Cover frontpage _gray

EUROPEAN SOCIETY FOR THERAPEUTIC RADIOL

Y AND ONCOL

OMINIQUE

UYSKENS

OGAERTS

ERSTRAETE

ARIKA

ÖÖF

ÅKAN

YSTRÖM

LAUDIO

IORINO

ARA

ROGGI

ÚRIA

ORNET

ONTSERRAT

IBAS

AVID

I. T

HWAITES

Sponsored by

“Europe Against Cancer”

RACTICAL

UIDELINES

MPLEMENTATION

IVO

OSIMETRY

ITH

IODES

XTERNAL

ADIOTHERAPY

ITH

HOTON

EAMS

NTRANCE

OSE

)

PHYSICS FOR CLINICAL RADIOTHERAPY

BOOKLET No. 5

Entrance in vivo dosimetry with diode detectors has been demonstrated to be a
valuable technique among the standard quality assurance methods used in a radio-
therapy department. Although its usefulness seems to be generally recognised, the
additional workload generated by in vivo dosimetry is one of the factors that
impedes a widespread implementation. Especially during the initial period of es-
tablishing the technique in clinical routine, the responsible QA person is con-
fronted with variable tasks, such as purchasing equipment, calibrating, defining
measurement and interpretation procedures. Often, this is accompanied by the
time-consuming activities of searching through literature and contacting expe-
rienced departments in order to gather information and define the sequence of the
steps to be undertaken.
This booklet is set up as a tool to reduce these initial efforts: it is conceived as a
step-by-step guide to implement entrance in vivo dosimetry with diodes in the
clinical routine of a radiotherapy department.
The first chapter about the preparation of the measurements contains information
(including commercial specifications) on diodes, electrometers and software.
Practical guidelines for the calibration of the diodes and the determination of cor-
rection factors are given.
The second chapter discusses the actual tasks of the responsible QA person dur-
ing the initial training period, with the emphasis on the implementation of the
measurement procedure (e.g. the training of personnel with explanation of imme-
diate actions to be undertaken in case of out-of-tolerance measurements)
In the third chapter, the interpretation of the measurement in relation to tolerance
and action levels is discussed and possible origins and consequences of out-of-tol-
erance measurements are given.
In an additional chapter, we present an overview resulting from the evaluation of
a questionnaire on how in vivo dosimetry has been implemented in different inter-
national centres. In the final chapter, elaborate contributions are given from five
centres about particular topics in in vivo dosimetry.

ISBN 90-804532-3

RACTICAL

UIDELINES

MPLEMENT

TION

IVO

OSIMETRY

ITH

IODES

XTERNAL

ADIOTHERAP

ITH

HOTON

EAMS

NTRANCE

OSE

)

TW KAFT BOOKLET 5 11-09-2001 11:46 Pagina 1

D.P. Huyskens, R. Bogaerts, J. Verstraete, M. Lööf, H. Nyström, C. Fiorino, S. Broggi,

N. Jornet, M. Ribas, D.I. Thwaites

Practical guidelines for the implementation of in vivo dosimetry with diodes

in external radiotherapy with photon beams (entrance dose)

2001 – First edition

ISBN 90-804532-3

No part of this publication may be reproduced,

stored in a retrieval system, or transmitted in any form or by any means,

electronic, mechanical, photocopying, recording or otherwise

without the prior permission of the copyright owners.

SUMMARY

Entrance in vivo dosimetry with diode detectors has been demonstrated to be a valuable

technique among the standard quality assurance methods used in a radiotherapy

department. Although its usefulness seems to be generally recognized, the additional

workload generated by in vivo dosimetry is one of the factors that impedes a widespread

implementation. Especially during the initial period of establishing the technique in clinical

routine, the responsible QA person is confronted with variable tasks, such as purchasing

equipment, calibrating, defining measurement and interpretatio n procedures. Often, this is

accompanied by the time -consuming activities of searching through literature and

contacting experienced departments in order to gather information and define the sequence

of the steps to be undertaken.

This booklet is set up as a tool to reduce these initial efforts: it is conceived as a step-by-

step guide to implement entrance in vivo dosimetry with diodes in the clinical routine of a

radiotherapy department.

The first chapter about the preparation of the measurements contains information

(including commercial specifications) on diodes, electrometers and software. Practical

guidelines for the calibration of the diodes and the determination of correction factors are

given.

The second chapter discusses the actual tasks of the res ponsible QA person during the

initial training period, with the emphasis on the implementation of the measurement

procedure (e.g. the training of personnel with explanation of immediate actions to be

undertaken in case of out-of-tolerance measurements)

In the third chapter, the interpretation of the measurement in relation to tolerance and

action levels is discussed and possible origins and consequences of out -of-tolerance

measurements are given.

In an additional chapter, we present an overview resulting from the evaluation of a

questionnaire on how in vivo dosimetry has been implemented in different international

TABLE OF CONTENT

Introduction ……………………………………………………………………………….10

Chapter 1 "Getting started” ........................................................................................................13

1.1

Equipment.......................................................................................................................13

1.1.1

Diodes ......................................................................................................................13

1.1.2

Electrometer.............................................................................................................16

1.1.3

Software ...................................................................................................................16

1.1.4

Commercially available equipment .......................................................................18

1.2

Calibration procedures ..................................................................................................22

1.2.1

Validation before use .............................................................................................22

1.2.2

Calibration of the diode for entrance dose measurements ...............................23

1.2.3

Determination of correction factors .....................................................................25

1.2.4

Long term performance ..........................................................................................28

Chapter 2 Implementation o f the measurement procedure in clinical practice ....................29

2.1

Training period: initial tasks of the Responsible QA person..................................29

2.2

Defining guidelines for the persons performing the measurements ......................31

2.3

Recording of in vivo dosimetry ...................................................................................35

Chapter 3 Interpretation of the measurement...........................................................................36

3.1

Defining tolerance and action levels ..........................................................................36

3.2

Which errors can be detected?....................................................................................39

3.2.1

Malfunctioning of the quality control process ..................................................39

3.2.2

Deviations in the treatment process (dosimetric errors)...................................41

3.2.2.1

Errors in data generation and data transfer (human errors) .................................................... 41

3.2.2.2

Errors due to equipment breakdown or malfunctioning ....................................................... 44

3.2.2.3

Discrepancies in patient positioning/geometry between treatment planning and delivery ........... 45

3.3

Evaluation of in vivo dosimetry data..........................................................................45

3.3.1

Actions after the first measurement.....................................................................45

3.3.2

Persisting deviations: interpretation of the result .............................................46

3.3.3

Monitoring in vivo dosimetry with time ..............................................................48

Chapter 4 Techniques and procedures in different radiotherapy centres ...........................50

4.1

What equipment do you use to carry out routin e in vivo dose measurements?.50

4.2

Philosophy of your department concerning the use of in vivo dosimetry? .........53

4.2.1

When do you use in vivo dosimetry? .................................................................53

4.2.2

What d o you measure?..........................................................................................54

4.3

Procedure for in vivo dosimetry? ................................................................................57

4.3.1

Calibration procedure? Which correction factors are used? ...........................57

4.3.2

Which measured and expected doses are compared? ......................................59

4.3.3

Value of tolerance and action levels + actions undertaken..............................61

4.3.4

Time period between checks of calibration and correction factors ................65

4.4

What system do you use for the recording of in vivo dose measurements?.......66

4.5

Workload? Specific tasks of people involved?.........................................................69

4.6

Examples of practical problems?..................................................................................73

Chapter 5 Experiences from different radiotherapy centres ...................................................76

5.1

Calibration and measurement procedures – The Barcelona experience ................76

5.1.1

Tests performed before diode calibration ...........................................................76

5.1.1.1

Signal stability after irradiation...................................................................................... 78

5.1.1.2

Intrinsic precision ....................................................................................................... 78

5.1.1.3

Study of the response/dose linearity................................................................................ 78

5.1.1.4

Verification of the water equivalent depth of the measuring point.......................................... 78

5.1.1.5

Perturbation of the radiation field behind the diode ............................................................ 79

5.1.2

Diode calibration (entrance dose)........................................................................79

5.1.2.1

Field size correction factor (CF

)................................................................................... 81

5.1.2.2

Tray correction factor (CF

tray

)........................................................................................ 83

5.1.2.3

Wedge correction factor (CF

wedge

)................................................................................... 83

5.1.2.4

SSD correction factor (CF

SSD

) ........................................................................................ 86

5.1.2.5

Angle correction factor (CF

angle

)..................................................................................... 87

5.1.2.6

Temperature correction factor (CF

temperature

) ...................................................................... 91

5.1.2.7

Influence of the dose rate on the diode’s sensitivity ........................................................... 93

5.1.2.8

sensitivity variation with accumulated dose (SVWAD) ...................................................... 94

5.2

Performance of some commercial diodes in high energy photon beams –

The Leuven experience .................................................................................................98

5.2.1

Introduction.............................................................................................................98

5.2.2

Material and Methods ...........................................................................................99

5.2.2.1

Material .................................................................................................................... 99

5.2.2.2

Methods...................................................................................................................102

5.2.3

Results ....................................................................................................................103

5.2.3.1

Independence of field size and SSD correction factors ........................................................103

5.2.3.2

Field size correction factor C

without tray.....................................................................104

5.2.3.3

SSD correction factor without tray.................................................................................106

5.2.3.4

Influence of beam modifiers: tray and block correction factor C

and C

................................108

5.2.3.5

Wedge correction factors..............................................................................................109

5.2.3.6

Correction factor variation within the same batch .............................................................110

5.2.3.7

Perturbation effects .....................................................................................................110

5.2.4

Discussion .............................................................................................................111

5.2.4.1

Independence of field size and source-to-surface distance correction factors .............................111

5.2.4.2

Total build-up thickness of the diode.............................................................................111

5.2.4.3

Treatment unit dependence ..........................................................................................112

5.2.4.4

Beam modifiers .........................................................................................................113

5.2.4.5

Perturbation effects .....................................................................................................113

5.2.5

Conclusion.............................................................................................................114

5.3

Practical implementation of cost-effective approaches to in vivo dosimetry -

The Edinburgh experience..........................................................................................115

5.3.1

Introduction...........................................................................................................115

5.3.2

Initial physics testing and workup.....................................................................117

5.3.3

Pilot clinical studies ..............................................................................................119

5.3.4

Routine use............................................................................................................122

5.3.5

Methods to simplify routine use........................................................................125

5.3.5.1

Possible omission of correction factors ...........................................................................125

5.3.5.2

The use of build-up caps .............................................................................................126

5.3.5.3

The use of ‘generic’ correction factors ............................................................................127

5.3.5.4

Data communication and recording................................................................................128

5.3.5.5

Diode mounting and handling ......................................................................................130

5.3.5.6

Diode quality control..................................................................................................131

5.4

Large scale in vivo dosimetry implementation –

The Copenhagen experience......................................................................................132

5.4.1

Introduction...........................................................................................................132

5.4.2

Methodology ........................................................................................................132

5.4.3

Equipment..............................................................................................................133

5.4.4

Calibration procedure ...........................................................................................133

5.4.5

Correction factors .................................................................................................135

5.4.6

Tolerance levels ....................................................................................................135

5.4.7

Results and discussion........................................................................................136

5.4.8

Conclusion.............................................................................................................139

5.5

Results of systematic in vivo entrance dosimetry –

The Milano (HSR) experience ....................................................................................141

5.5.1

Materials and methods ........................................................................................141

5.5.1.1

Equipment................................................................................................................141

5.5.1.2

In vivo measured and expected entrance dose...................................................................141

5.5.1.3

QA chain: methods ....................................................................................................142

5.5.1.4

MU calculation/data transfer check.................................................................................144

5.5.2

Results ....................................................................................................................144

5.5.2.1

Detection of systematic errors.......................................................................................144

5.5.2.2

Systematic errors detected before in vivo dosimetry by MU calculation/data transfer check .......145

5.5.2.3

Patients with more than one check ................................................................................145

5.5.2.4

Accuracy of treatment delivery ......................................................................................145

5.5.3

Final remarks..........................................................................................................149

Appendix 1 : Literature overview.......................................................................................140

INTRODUCTION

The aim of this booklet is to provide the radiotherapy community with practical guidelines

for the implementation of in vivo dosimetry (IVD) with diodes at a routin e/departmental

level.

Since in vivo dosimetry with diodes is a broad subject, considering the full map of varieties

encountered in radiotherapy, the authors have restricted themselves to guidelines for the

measurements of the entrance dose with diodes in p hoton beams. This technique of in vivo

dosimetry is the first to be considered by a radiotherapy department planning to start with

in vivo dosimetry as a routine QA method. As such, entrance diode measurements

supplement and complement basic pre -treatment QA methods, such as the independent

check of dose calculation and data transfer, which should be in routine use in the

department prior to the implementation of in vivo dosimetry.

The information contained in this booklet is a practically usable distillate from other

publications on in vivo dosimetry. The literature overview is set up as a database, which

includes, for the sake of completeness, publications dealing with exit dose measurements

and midline dose calculations, and in vivo dose measurements in electron beams. The way

in which the information is presented has rendered a booklet that is complementary to

other ESTRO booklets on in vivo dosimetry - that have appeared and will appear – and to

other review publications. These include the first ESTRO booklet on in vivo dosimetry

“Methods for in vivo dosimetry in external radiotherapy”, written by J. Van Dam and G.

Marinello

[

Van Dam 1994

]

, and a new ESTRO booklet “In vivo dosimetry in clinical

practice: When and What to measure? How to correct?”, written by E. van der Schueren,

A. Dutreix and C. Weltens [van der Schueren 2001]. A nice general review on in vivo

dosimetry was written by M. Essers and B. Mijnheer

[

Essers 1999

]

The latter two publications highlight the more philosophical questions concerning the use

of in vivo dosimetry. These will not be discussed here. Also, the future use of diode

measurements in relation to conformal irradiation techniques and IMRT - for which point

dose verification is obviously inadequate – is a topic outside the scope of this work.

CHAPTER 1 "GETTING STARTED”

1.1 EQUIPMENT

1.1.1 DIODES

Semiconductor diodes, when connected to a suitable electrometer, offer the unique

combination of high sensitivity, immediate readout, simplicity of operation (no external bias

voltage), small size and robustness.

Silicon diodes can be made starting from either n-type or p-type silicon, which behave

differently because their minority carriers are holes or electrons, respectively. Figure 1.1

illustrates the basic operation of a p-type silicon detector diode. In the boundary between

two regions, one of p-type and another of n -type silicon, there is a depletion of free charge

carriers. When the detector is operating with zero external voltage a potential difference of

about 0.7 V exists over this depletion area, causing the charge carriers created by the

radiation to be swept away into the body of the crystal. As the diode is asymmetrically

doped - the n-type region is much more heavily doped than the p-type region - the

irradiation induced charge flow is comprised almost entirely of electrons (holes in an n -type

diode). Due to defects in the crystal lattice some electrons are trapped and will

consequently not contribute to the diode signal. An n-type diode is more influenced by

these recombinations as holes are more easily trapped than electrons.

Figure 1.1 Schematic overview of the basic principal of a p-type silicon diode used

as a radiation detector.

depletion layer

p-type

+ +

+
+

+
+ +
+

-
-

-
- +

n-type

electrostatic potential ~ 0.7 V

The detector sensit ivity depends on the lifetime of the charge carriers and consequently on

the amount of recombination centres in the crystal, which is determined by the diode type,

the doping level and the accumulated dose. As the radiation induces recombination

centres within the lattice, the sensitivity will decrease with accumulated dose.

Figure 1.2 Schematic summary of the factors (physical as well as geometrical)

influencing the diode signal. Arrows indicate dependencies of one factor on

another. The different influences are taken into account in calibration and

correction factors (see Section 1.2).

The effect of radiation damage represents the main limitation of silicon diodes.

Furthermore, other effects related to the detector material have to be considered (Figure

1.2):

•

The diode signal depends on the photon energy. This is due to the higher atomic

number of silicon (Z = 14) compared to soft tissue (Z =

∼

7) and the corresponding

higher contribution to the diode signal from the photo-electric effect.

•

The diode signal is dose rate dependent. At high instantaneous dose rates the

recombination centres will be “occupied” resulting in a rela tively lower rate of

recombination. This leads to a proportionally higher response at higher dose rates. This

effect is more pronounced for n-type Si-diode detectors than for those made of highly

CALIBRATION

CORRECTION FACTORS

physical dependencies

geometrical dependencies

related to detector material

+ related to measurement methodology

+ related to build-up cap design/thickness

energy

field size

accumulated dose

SSD

tray, blocks

temperature

wedge

dose rate

orientation

doped p-type Si [Heukelom 1991b]. The dose rate dependence may change with

accumulated dose due to radiation damage.

•

The diode signal is influenced by temperature. In general, the sensitivity increases with

increasing temperature. This effect is less pronounced for an un-irradiated diode and

will increase with accumulated dose. However, the rate of change of sensitivity with

temperature tends to stabilise as accumulated dose increases [Grusell 1986].

Due to these dependencies, the “true” entrance dose dependence on geometric

parameters, as SSD (source skin distance), field size, and the presence of wedges, trays and

blocks, will be incorrectly reflected by the diode signal variation. For this reason, correction

factors have to be determined (see Section 1.2.3 describing t he practical details). Beside the

physical properties of the diode crystal, other factors contribute to the magnitude of these

correction factors (Figure 1.2). First, there is the inevitable fact that the measureme nts are

performed with the diode located outside the patient or the phantom. The photon scatter

conditions experienced by the diode are therefore different from those at the point of

entrance dose definition, i.e. at the depth of dose maximum inside the patient or the

phantom. For instance in high energy beams the diode reading is nearly independent of

phantom scatter, while the entrance dose is clearly not (see [Jornet 2000], [Lööf 2001],

[Wierzbicki 1998]). In addition, the diode may experience a different amount of head-scatter

electrons. As a consequence, the effect of field size, SSD, and presence of wedges, trays

and blocks on the contaminating head-scatter electron contribution may induce large

variations in the diode correction factors (see Section 5.2).

In relation to this, the construction of the detector, for instance the thickness and the

shape of the build -up cap, is another factor influencing the diode signal. The shape of the

build-up cap will influence the angular response: a cylindrical cap has a different angular

dependence than a hemi -spherical one. The thickness of the build -up cap determines the

scatter conditions seen by the diode. To minimise the correction factors and ensure a

greater accuracy in the measurements, it is preferable to have a (possibly in house made)

build-up cap with a thickness equal to the depth of maximum dose (see Sections 5.2 and

5.3). On the other hand, it should be kept in mind that a thick build -up cap means a larger

perturbation of the treatment field and may jeopardise the dose to the patient if

measurements are performed during many fractions of the treatment course. In addition, if

the same diode is used for different beam qualities (for instance 6 MV and 18 MV), it may

be preferable to use the same build -up cap for both, in order to avoid confusion and

interchange of build -up caps. It follows therefore that the choice of the "optimal" diode

design is a question of each department’s policy for the in vivo dosimetry procedure.

1.1.2 ELECTROMETER

The diode should be connected to a dedicated electrometer with a low input impedance

and low offset voltage. Diode current generated by sources others than radiation is

considered to be leakage current and is not desirable. The leakage current ideally should be

zero. Due to the input offset voltage of the amplifier, however, there is always a small bias

across the diode introducing a small leakage current. An electrometer used together with a

diode requires therefore the offset voltage of the amplifier to be low, 10

V or less. The

leakage current increases with temperature and accumulated dose due to defects in the

diode and it is essential that the electrometer has adequate zero drift compensation and

stabilisation.

1.1.3 SOFTWARE

There is a range of electrometers available for in vivo dosimetry, having greater or lesser

degree of sophistication. The simplest type of electrometer provides 5 to 10 chann els with

manual adjustment of the input offset and gain for each channel. This type of electrometer

may allow only one gain setting for each channel while more sophisticated ones offer

several separate calibration sets and correction sets with automatic calculation, storage of

factors and zero drift compensations. Thus one detector may be calibrated to be used in

several different irradiation conditions. Most of the electrometers offer the possibility to

use interface software designed to run in a Windows environment in conjunction with

commercial available software or in house made programs loaded onto a personal

computer.

More advanced systems are incorporated with the department’s verification system,

simplifying the management of the in vivo dosimetry procedure. Such system provides the

possibility to store all calibration and correction factors for every diode in use. The

measured diode signal is then automatically converted to dose using the treatment field

parameters downloaded from the patient’s data in the verification system. This gives an

immediate "on line" check of the preparation and treatment delivery in the radiotherapy

process, thereby reducing the incidence of errors.

CONSIDERATIONS WHEN CHOOSING EQUIPMENT

•

Pre-irradiated diodes have in general lower sensitivity. This parameter has to be taken

into account when choosing the electrometer, which must have a sensitivity range that

matches the diode. The manufacturer of the detectors usually also supplies adequate

electrometers.

•

The diodes are available in negative and positive polarity and the electrometer has to

be adapted to this.

•

The rate of sensitivity degradation will affect the required calibration frequency. In

general, n-type diodes have larger sensitivity degradation but the rate of degra dation

will decrease after a certain amount of pre -irradiation (both for n- a n d p-type).

Therefore, the pre -irradiation level is of interest and should be stated by the

manufacturer.

•

The best choice of diode design i.e. shape and thickness of the build -up cap depends

on the application. A cylindrical cap (uniform directional response around the detector

axis) is preferable in measurements in tangential treatment techniques while in entrance

dose measurements with perpendicular incident beams a hemi -spherical is a better

choice (smaller perturbation than the cylindrical cap). If little influence on the

perturbation of the treatment field is desired (measurements in all sessions) a diode with

thin build-up cap is preferable. This however means that larger correction factors have

to be used for accurate measurements, whereas diodes with thicker caps need smaller

correction factors at the expense of larger perturbation.

•

A number of properties of importance for the clinical use of diodes are related to their

dose rate dependence. If diodes are to be used in irradiation situations with large

variations of dose per pulse i.e. wedged fields or treatments at SSD deviating from

calibration SSD, it is advisable to choose diodes with a low dose per pulse dependence,

usually high doped p -type diode detectors.

1.1.4 COMMERCIALLY AVAILABLE EQUIPMENT

There are several different types of diodes commercially available having various

properties with regard to pre -irradiation level, doping type, design and thickness of build -

up cap to accommodate a large photon energy range. For accurate in vivo dosimetry it is

essential that each diode characteristic is well understood in order to utilise it properly and

efficiently. Unfortunately though, manufacturer’s specifications can sometimes be difficult

to interpret. Some specifications should be handled with caution: for instance, it is known

that several commercial diodes lack sufficient build -up for the energy range that they are

specified for (see [Jornet 1996], [Georg 1999] and [Meijer 2001]).

Table 1.1 to Table 1.6 present diodes and selected specifications available from leading

companies.

Company

Scanditronix

Medical AB

EDP diodes

Sun Nuclear

Corporation

QED diodes

Precitron MDS

Nordion AB

P diodes

Nuclear

Associates

VeriDose diodes

Sun Nuclear

Corporation

Isorad-p diodes

Type

p-type

n-type

p-type

Sensitivity

[nC/Gy]

150-300

150

Sensitivity

degradation

[% / kGy]

< 1.5 - 10 MeV

< 1.5 - 18 MV

0.1 at 6 MV

< 15 %

after 10 kGy

1 at 10 MeV

0.1 at 6 MV

Sensitivity

degradation

with

temperature

[% /

0.4

0.3

0.1 - 0.3

< 0.5

0.3

Pre irradiation

level

8 kGy at 10

MeV

10 kGy at 10

MeV

25 kGy

10 kGy at 10

MeV

Linearity

(Dose per

pulse

dependence)

< 1%

in the range

0.1 - 0.6 mGy

per pulse

SSD

dependence

2% for

18MV

1% for 8MV

at SSD 80-130

cm for a

typical

accelerator

< 1% of SSD

1% for

6 and 18MV

at SSD 80-130

cm for a

typical

accelerator

Output

polarity

Negative

Negative or

positive

Negative

Negative or

positive

Negative or

positive

Table 1.1 Commercially available diodes listed along with the specifications given by

the manufacturers.

Model

Application /

Energy range

Build-up cap /

Water equivalent build -

5 mm

P10

4 - 8 MV

10 mm

P20

8 -16 MV

20 mm

P30

16 - 22 MV

tungsten / 30 mm

Table 1.2 Precitron–Helax, P-diodes

Model

Application /

Energy range

Build-up cap /

Water equivalent build -

EDD-2

Entrance / Exit

Electrons

paint / 2mm

EDD-5

Risk organ

monitoring

polystyrene / 4.5 mm

EDP- 0

Skin dose

None

EDP-10

4 - 8 MV

stainless steel / 10 mm

EDP-15

6 -12 MV

stainless steel / 15 mm

EDP-20

8 -16 MV

stainless steel / 20 mm

EDP-30

16 - 25 MV

tantalum / 30 mm*

*less than 30 mm [Jornet 1996], [Meijer 2001]

Table 1.3 Scanditronix Medical AB, EDP diodes

Model

Application /

Energy range

Build-up cap /

Water equivalent build -

1112

Electron

acrylic

1113

Skin dose /

Scatter dose

none

1114

1 -4 MV

aluminium / 10 mm

1115

6 -12 MV

brass / 18.5 mm

1116

15 -25 MV

brass / 30.4 mm

Table 1.4 Sun Nuclear Corporation, QED diodes

Model

Application /

Energy range

Build-up cap /

Water equivalent build -

30-475

6 – 25 MeV

30-471

1 – 4 MV

copper / 7 mm

30-472

5 – 11 MV

copper / 14 mm

30-473

12 –17 MV

tungsten / 26 mm

30-474

18 - 25 MV

tungsten / 36 mm

Table 1.5 Nuclear Associates, VeriDose diodes

Model

Application /

Energy range

Build-up cap /

Water equivalent build -

1162

1 - 4 MV

aluminium / 10 mm

1163

6 -12 MV

brass / 18.5 mm

1164

15 - 25 MV

tungsten / 30.4 mm

Table 1.6 Sun Nuclear, Isorad-p diodes.

1.2 CALIBRATION PROCEDUR ES

1.2.1 VALIDATION BEFORE US E

The signal stability of the diode, influenced e.g. by the leakage current without irradiation,

should be checked after adequate warm-up time with the diode connected to the

electrometer and compensated. Compared to the current obtained for the real measurement,

the leakage current should be insignificant. It is advisable to measure the leakage curr ent

for a time period that is at least five times longer than the time period used in the clinical

application. The leakage current should not exceed 1% in one hour

[

Van Dam 1994

]

A general test of the reliability and stability of the equipment, before using it in clinical

routine, can be performed as follows. The diode positioned on top of a calibration phantom

(see Section 1.2.2) is irradiated for 10 to 15 times with the same reference field. The standard

deviation of the resulting signals should be within 0.5 %. The measurements are repeated

on different days during two weeks. The measurement procedure, including the

measurement equipment, the phantom set-up and diode positioning, is reliable and stable,

if all measurements are within 1 % (provided that the beam output of the treatment unit is

stable).

Some centres perform more extensive tests before using the diode, for instance a

measurement of the effective water equivalent thickness of the build -up cap. An example of

this can be found in Section 5.1.

1.2.2 CALIBRATION OF THE DIODE FOR ENTRANCE DOSE MEASUREMENTS

The diode is calibrated to measure the entrance dose, i.e. when positioned on the skin of

the patient the measured dose should correspond to the dose to tissue at the depth of

maximum dose of the photon quality in use for a particular beam geometry.

The calibration procedure firstly involves the determination of the calibration factor (F

cal

It is recommended to calibrate the diode for each beam quality with which it is intended to

be used (see Figure 1.2). Due to the variation of the diode signal with accumulated dose,

calibration should be regularly repeated in time. Time intervals typically vary between

weekly and monthly. The temperature dependence of the diode signal can be accounted for

during calibration, if this is performed at the same temperature as the measurements with

that particular diode in the clinical application. Usually, however, a temperature correction

factor will be determined (see Section 1.2.3).

The entrance dose value in a clinical situation is calculated from the diode measurement as

the product of the diode reading, the calibration factor and the correction factors (equation

1). The calibration factor is defined as the ratio of the ion chamber dose and the diode

reading measured in the reference geometry (equation 2).

∏

⋅

cal

diode

entr

(1)

condition

ref

diode

cal









(2)

determination with the ionisation chamber includes the application of a displacement

factor, this factor should be omitted. If a plastic phantom is used that is not completely

water-equivelant (for instance made of polystyrene), a conversion factor dose to p lastic –

dose-to-water should be employed. The reference SSD is usually 100 cm (for linacs) and

the reference field 10 x 10 cm

The calibration may be performed with one or several diodes placed in a circle around the

central axis provided that variations in the field flatness are insignificant. Field flatness at

max

should therefore be checked, for instance by measuring the ratio of diode readings at

the circle and at the centre of the field. Furthermore, the diodes should be placed at a

distance from the central axis to avoid perturbation of the beam at the reference chamber.

1.2.3 DETERMINATION OF CORRECTION FACTORS

Subsequent to the determination of the calibration factor, a set of correction factors has to

be established to account for variations in diode response in situations deviating from the

reference conditions (see Figure 1.2). The ultimate factors influencing the diode response

are the field size, source-to-skin distance, presence of beam modifiers such as filters or

wedges, presence of tray and blocks and the beam incident angle (equation 3). As

described in Section 1.1.1, the dependence of the diode signal on most of these factors is

not only arising from the intrinsic properties of the diode crystal, but also from elementary

beam physics, i.e. the fact that the detector experiences different scatter contributions than

the ones experienced at the depth of maximum dose. As a consequence, most of the

correction factors are inherent to the use of dose detectors taped to the patient’s skin, and

should also be applied e.g. for thermoluminescent dosimeters (TLDs).

The temperature dependence, intrinsic to diodes, should be accounted if a particular diode

is used at different temperatures. This may be done by applying a constant temperature

correction factor or by using a thermostatically controlled calibration phantom. However, if

the patient measurement is assessed before the diode has reached thermal equilibrium (2 -3

minutes) the influence of temperature dependence may be neglected (see Sections 5.3.3

and 5.5.1). Diodes used for TBI, for which the low dose rate is achieved by enlarging the

SSD, should be calibrated in TBI conditions.

Correction factors accounting for the variations in response are determined as the ratio of

the reading of an ionisation chamber and the reading of the diode for a clinical irradiation

situation normalised to the same ratio for the refe rence situation (equation 4):

angle

block

tray

wedge

SSD

(3)

condition

ref

diode

condition

clinical

diode

)

(

)

(

(4)

The reference conditions are as stated in Section 1.2.2. However, if the value of a particular

parameter (for instance the field size) influences the value of the correction factor for a

second parameter (e.g. the presence of a tray), the ‘reference condition’ for the

determination of this second correction factor is adapted in order to avoid double inclusion

of the first correction factor. This is made clear in the practical recommendations given

below. These may be modified/simplified according to the type of diode (and previous

experience with that particular type of diode), the clinical application, and the beam quality

in use:

•

The variation in response due to different beam incident angles is measured for

different gantry and couch angles and normalised to the response measured when the

central beam axis and the symmetry axis coincide.

•

Field size correction factors are measured for square fields ranging e.g. from 5 x 5 cm

40 x 40 cm

, at the reference SSD of 100 cm.

•

SSD correction factors are measured for SSDs within a range determined by local

clinical conditions, for instance from 75 cm to 110 cm, at the reference field of 10 x 10

. Note that SSD correction factors and field size correction factors are assumed to be

independent. This is not always the case for high energies (see remark below and

[Georg 1999]).

•

Wedge correction factors may depend on the field size. They are measured at reference

SSD, for different square fields (e.g. fields with 5 cm, 10 cm and 20 cm side length). The

ratio of the signal of the ionisation chamber to the diode signal is in this case

normalised to the same ratio for the open beam (with the same field size).

•

Tray correction factors may depend on SSD and field size. They can be determined by

repeating all measurements carried out for the SSD and field size correction factors, and

normalising the data to the reference situation of an open beam with the appropriate

SSD and field size.

•

Block correction factors can be measured for different blocks defining square fields at a

fixed collimator opening (for instance a collimator opening of 20 x 20 cm

for blocks

defining fields of 5 x 5 cm

, 10 x 10 cm

and 15 x 15 cm

). The reference condition is again

the corresponding open beam (with the same collimator opening).

Practical examples of calibration procedures and typical values of correctio n factors for

particular types of diodes are given in Section 5.1. In order to minimise redundant use of

correction factors, minimum values can be set below which factors are discarded (for

instance if a correction factor deviates less than 1 % from 1, the correction is within the

measurement uncertainty). Other ways of limiting the use of correction factors are

described in Section 5.3.5.

If the dosimetric characteristics of a diode are not (well) known, it is recommended to check

its response extensively at different irradiation conditions to establish the range where no

correction factors are needed. As the type of diode is a major determinant of the magnitude

and the behaviour of most of the correction factors, diodes of the same type will require

similar correction factors, showing similar tendencies. However, when a high accuracy is

required, it is advisable to check also the correction factors for every individua l diode.

Correction factors associated with increased diode sensitivity due to variation in beam

energy spectrum are of major importance in high-energy photon beams, especially if diodes

with a thin build -up cap are used. One should also bear in mind that, for insufficient build -

up, other interdependencies of correction factors than the ones mentioned above may exist

[Georg 1999]. In practice, one can start by considering all correction factors to be

independent and then check the accuracy of the measured d ose when changing more than

one reference condition at the same time (i.e. field size, SSD and wedge). Useful information

about the performance of diodes in high-energy beams is given in Section 5.2.

1.2.4 LONG TERM PERFORMANCE

It is good practice to keep a record of the change in the calibration factor in order to

estimate how often re -calibration will be required to achieve a certain accuracy. As the

sensitivity degradation may vary with different beam qualities this is especially important

when diodes are used in various beam qualities. It is advisable to start with weekly

calibrations and to adjust the calibration interval after having monitored the accumulated

dose in between calibrations and the corresponding change in calibration factor for a while.

Depending on the diode type in use the correction factors associated with the dose per

pulse dependence may also change with time due to the accumulated dose. A quick and

efficient test of the long-term stability is to perform a linearity check by measuring the

diode response normalised to an ionisation chamber at two different SSDs. If the ratio is as

expected, the diode is working accurately and the correction factors are still valid. The

change of the temperature dependence with time is accounted for if the diode is calibrated

at the same temperature as the measurements in the clinical application.

CHAPTER 2 IMPLEMENTATION OF THE MEASUREMENT

PROCEDURE IN CLINICAL PRACTICE

Regarding the workload associated with routine in viv o dose measurements, two

categories of work can be distinguished: the calibration procedures and the actual patient

measurement procedures. Depending on the strategy of the radiotherapy department, these

procedures can either be carried out by a small team of qualified personnel, or assigned to

different groups of personnel within the department (for instance calibration procedures

are carried out by a dosimetrist/physicist; patient measurements are performed by the

radiographers/nurses at the treatment units). In the latter case, one person or a small group

of persons have the responsibility for the in vivo dosimetry program and train the others.

2.1 TRAINING PERIOD: INITIAL TASKS OF THE RESPONSIBLE QA PERSON

The person(s) responsible for the in vivo dosimetry program initiates the implementation of

it. First, he/she gets acquainted with the theoretical background of in vivo dosimetry,

available in literature (see appendix), with the equipment used in the department

(electrometer and diodes) and with the calibration techniques using the (plastic) calibration

phantom. The test of the reliability and stability of the equipment (see Section 1.2.1) is

performed. After the reliability test, the electrometer is calibrated following the instructions

in the manual, and the calibration and correction factors of the diodes are determined as

prescribed in Section 1.2.

Another task is the practical training for the personnel performing the measurements. The

importance of the accurate positioning of the diode in the centre of the treatment field is

emphasised. A demonstration is given by performing 10 irradiations with a wedge, e.g. 30°,

for which the positioning of the diode is critical. Between each irradiation the diode is

removed and repositioned. The readings of the electrometer should be within 1.5 %.

If the beam axis of the treatment field is covered by a shielding block or in case of an

asymmetric field, the penumbra region should be avoided by positioning the diode as close

as possible to the field centre and at a similar SSD. If it is a wedged field, the actual

attenuation of the wedge at the off-axis position of the diode should also be considered.

Before starting patient measurements, it is useful to simulate some patient set-ups with a

phantom. The irradiations and diode measurements are performed in identical conditions as

in the clinical situation. The expected signal is calculated, either with an independent

formula or with a treatment planning system (TPS) able to calculate dose at d

max

. The

difference between the calculated and the measured signal should not exceed 1 %. These

patient simulations are an excellent test for the whole measuring procedure: calibration of

the dio de and determination of correction factors, calculation of the expected dose, and

diode positioning.

Patient measurements should be started for treatment fields where easy fixation of the

diode in the field centre is possible, for instance mediastinal or large head and neck fields

without wedge. When the deviations between measured and expected signal are smaller

than 3% to 5%, measurements for other treatments like breast or pelvic irradiations can be

initiated.

In the course of the training period, tolerance and/or action levels have to be established.

Since a measurement result out of the tolerance window triggers the chain of measurement

interpretation, determination of the values of these levels is discussed in more detail as a

first item in Chapter 3. The adequacy of tolerance/action levels should be examined

regularly (see Section 3.1), and especially during the training period. In this period it is also

useful to keep track of other parameters:

•

the precision of a single measurement: this can be done by performing repetitive

measurements on the same patient during consecutive sessions (at least 5). The mean

value, the standard deviation and the deviation of each individual measurement are

evaluated. A small standard deviation is a strong argument for considering the value of

the first measurement as being representative for the whole treatment. This evaluation

should be made for different groups of patients and t reatments.

•

the calibration and correction factors (see Section 1.2.4).

2.2 DEFINING GUIDELINES FOR THE PERSONS PERFORMING THE

MEASUREMENTS

It is essential to define departmental guidelines and/or procedures describing

the

immediate actions to be taken when the measured entrance dose is out of the tolerance

and/or action levels. These guidelines will differ among the radiotherapy departments

depending on the choice of the general philosophy for in vivo dosimetry (which patients,

which fraction, which treatment sites etc.), the education level of radiation technologists,

the existence of a Quality Assurance group and/or the involvement of the physics

department.

Practical guidelines towards the radiation technologists, assuming that they are in charge

of these in vivo measurements, should provide an answer to the following questions:

•

if the measured entrance dose exceeds the tolerance or action levels, what

should be done (perform a new measurement, call the QA group/physicist, ...)?

•

if there is a difference between the stated SSD and the actual SDD (source-diode

distance), what should be done i) for isocentric techniques or ii) when using bolus or

immobilisation devices (correct for inverse square law, ... )?

Other questions regarding staffing and management of personnel should also be clarified:

•

who is the contact person for measurements out of tolerance or action levels?

•

if a second measurement is requested, should it be performed in the presence of a

physicist?

•

who will perform phantom measurements, if needed?

•

who is in charge of the calibration and the determination of correction factors of the

diodes?

Examples of guideline flowcharts, including actions undertaken at different levels, are given

in Figure 2.1 and Figure 2.2; more examples are given in the questionnaire of Chapter 4. The

possible origin of errors and the actions undertaken are discussed in more detail in

Sections 3.2 and 3.3. Typically, the investigation of an error is performed in two steps.

Because of the on-line read-out, the first action can be triggered instantaneously by the

technical staff, who performs an immediate check on the spot. If the origin of the error is

not found, an “a posteriori” check should be performed by a physicist/QA personnel.

out-of-tolerance

signal for most of the

patients that day?

immediate check:

patient set-up/ diode

error on IVD

chart?

phantom simulation

error on

dosimetry chart?

check diode

calibration factor

“a posteriori” check:

IVD chart check;

data transfer, MU check

is treatment OK?

yes

START:

entrance measurement

signal

within 5 %?

clearly wrong

diode/patient

position?

yes

error found?

correct /

recalibrate

check accelerator

output

error found?

yes

stop

treatments

stop treatments;

discussion

correct

error in MU

calculation?

stop treatment;

discussion

check diode

correction factors

yes

Figure 2.2 Example of a flowchart (taken from Leuven), guiding the actions to be

undertaken after an in vivo entrance dose measurement. The tolerance level

coincides with the lowest action level.

wedged field

or IMN (critical

diode position)?

signal

within tolerance?

immediate check “on the spot”:

record SSD and SDD;

check diode position

treatment

plan/data transfer

error?

new measurement
by QA personnel

phantom simulation

wrong SDD,

wrong calculated

signal?

signal within

action level 2?

second measurement

“a posteriori” check:
data transfer check;

recalculation expected signal

signal within

tolerance?

treatment plan

modification

yes

discussion with Head of

Physics Department;

check of equipment

START:

entrance measurement

signal

within tolerance

(= action level 1)?

clearly wrong

diode/patient

position?

yes

CHAPTER 3 INTERPRETATION OF THE MEASUREMENT

3.1 DEFINING TOLERANCE A ND ACTION LEVELS

The choice of tolerance/action levels is very important since they will in practice determine

the number of "errors" detected and will influence the associated workload t o implement or

to maintain in vivo entrance dose measurements at a departmental level. If a too broad

tolerance window is adopted, some causes of erroneous treatment delivery may not be

detected (for instance a wedge 30° instead of a wedge 15°, presence of a tray etc.). If the

tolerance window is too small, a too large number of measurements will have to be repeated

(due to e.g. inherent statistical fluctuation or a too critical positioning of the diode in e.g.

wedged beams). Clearly, the value and the meaning of the levels are related to the

philosophy of the department regarding in vivo dosimetry. Some centres using in vivo

dosimetry as a routine check for every patient, distinguish the first level, the tolerance

level, from higher action levels. A deviation of the diode signal beyond the tolerance level,

but within the action levels, is considered as a warning, linked to a very limited action.

When in vivo dosimetry is used to check particular treatments, the value of the levels can

vary according to the treatment type. Treatments with high dose - high precision

techniques require narrow tolerance windows, while other treatments have less stringent

accuracy demands. In certain centres, it could be realistic to set higher tolerance/action

levels for patients treated with palliative intent in order to minimise the number of second

measurements, paying more attention to the patients treated with curative or adjuvant

intent.

The determination of the actual value of the level is based on different factors, first of all on

the uncertainty of the diode measurement method. According to Essers and Mijnheer

[

Essers 1999

]

, the theoretical uncertainty in measuring the entrance dose with diodes,

taking into consideration the uncertainties in the calibration factor, the correction factors

and the positioning of the diode, is 1.6 % (1 standard deviation (SD)). This means that

without any additional cause for deviation or error, 68% of the measurements would be

within 1.6 % and 95 % of the measurements would be within 3.2 % (2 SD) of the expected

dose. This seems to be in agreement with other results reported in the literature, although

such a level of accuracy is probably difficult to obtain for all types of irradiation, for

instance for treatments with tangential wedged beams. It also has to be stressed that this

high level of accuracy is attainable only if a very accurate estimate of diode correction

factors is accomplished. It has to be kept in mind that a choice of “minimum” correction

factors, which could be preferable in small and medium centres with a small physics staff,

means a larger uncertainty in dose estimation and, consequently, results in the necessity of

higher tolerance/action levels.

Other sources of uncertainty, which should be taken into account when choosing the

levels, are:

•

the physiological movements due to breathing and/or possible movements of the

patient during irradiation; the difficulty in firmly attaching the diode in some regions

due to the presence of hair

•

the use of ancillary equipment to set-up the patient (i.e.: head masks, head-rest,

immobilisation shells…) [Essers 1994]. In these cases the diode reading has to be

corrected to take the real SDD into account, which can be difficult to assess in some

situations. Moreover, some loss of backscatter may occur in many situations, which is

another source of uncertainty, as this is usually not taken into account by the TPS.

Also, it should be kept in mind that the positioning of the diode on immobilisation

shells (or on the back of the couch when treating with dorsal fields) results in a larger

uncertainty if the temperature dependence of the diode signal is accounted for in the

calibration factor (see Section 1.2.2)

•

the true SSD, if the diode reading is not corrected to take into account the difference

between the true SSD and the planned one. If the correction is performed, in some

situations, it is difficult to calculate the correction factor (for instance in posterior-

anterior fields or in some tangential fields)

•

the use of asymmetric fields, e.g. for tangential breast treatments (see Section 5.4.2)

•

possible fluctuations of accelerator output

As the in vivo measured entrance dose has to be compared with the expected one, which is

calculated by the TPS or by an independent formula, the uncertainty in the entrance dose

calculation is another factor that should be taken into account. This uncertainty depends

on:

•

the algorithms used for dose calculation

•

the method used for calculating MUs

•

the way in which inhomogeneities are taken into account

•

the way in which treatment unit data have been acquired (for instance the precision

with

which d

max

has been determined)

The majority of the radiotherapy centres have a 5 % action level for most treatments (see

Chapter 4 and Sections 5.4). The tolerance level usually coincides with this level, according

to the philosophy that any deviation larger than 5 % must be investigated. A procedure

recommended for establishing the narrow tolerance window required for high dose - high

precision techniques is to investigate the attainable measurement accuracy for the

particular technique and take twice the SD of the measurement uncertainty as the

tolerance/action level ([Essers 1993], [Essers 1994], [Lanson 1999]). In this case, however,

entrance dose measurements are usually combined with exit dose measurements to obtain

the target absorbed dose.

Once the tolerance and action levels have been established, the range of acceptable

variation of some of the parameters can be determined in order to facilitate the search for

the cause of an out-of-tolerance signal. Acceptable deviations in stated SSD versus

measured SSD (or SDD) for isocentric and fixed SSD techniques can be determined. The

importance of daily beam output variations can be assessed.

It is important to verify during a certain period whether the tolerance/action levels are

adequate for clinical routine. An important indicator is the rate of second measureme nts,

which is strictly related to the action level. A too small rate (for instance less than 2-3%)

should be regarded with caution because it might indicate that the action level is too high.

Inversely a too high second check rate (for instance larger than 15-20 %) could mean that

the level is too small. In

particular a high rate of second checks can generate distrust

concerning the real usefulness of in vivo dosimetry among the operators and the medical

staff. An alternative method for adjusting the tolera nce/action levels is to adapt it to the SD

of the measurements. This parameter can be determined by pooling the patients for a

certain period.

It is clear that the continuous monitoring of systematic in vivo dosimetry after its

implementation is mandatory in order to reduce the errors of the control process and

possibly to adjust tolerance/action levels in time if they appear to be inadequate (see

Section 3.3.3). Such a monitoring could also help defining differe nt tolerance/action levels

for different types of patient treatments and/or beam set-up. For instance, it could become

clear whether wedged beams need higher tolerance/action levels with respect to unwedged

beams because of the corresponding larger uncertainty.

3.2 WHICH ERRORS CAN BE DETECTED?

It is important to keep in mind that when a deviation is observed out of the tolerance level,

it is not necessarily an error in the treatment process but it may be linked to a

malfunctioning of the quality control process.

3.2.1 MALFUNCTIONING OF THE QUALITY CONTROL PROCESS

Deviations between measured and prescribed entrance dose due to an erroneous

measuring procedure at a departmental level can affect the confidence in in vivo dosimetry.

If, for instance, during a chart round a radiation oncologist finds out that most of the in

vivo measurements are out of tolerance (due to a problem in the QC process), it is rather

difficult to yet convince him that his patients are indeed correctly treated, and/or that in

vivo measurements are useful.

Malfunctioning of the QC process (Figure 3.1), may be present either at the departmental

level, leading to systematic errors (i.e. for all patients), or at the individual level. Systematic

errors are typically errors (or a shift) in the calibration factors of the diodes or an error in

the correction factors (or omitting some necessary correction factors). Systematic errors

may also be induced by erroneous calculation (with or without TPS) of the entranc e dose.

Depending on the procedures of the department, systematic deviations in the QC process

may increase the workload since one might start to look for a “real” dosimetric error and/or

one might request a new measurement to exclude other “individual” errors in the QC chain.

The individual

errors in the QC chain for entrance dose are mainly the following ones:

•

miscalculation of the expected diode signal from the entrance dose (use of wrong

calibration factor, correction factors), which irritates personnel

•

misrecording the SDD

•

erroneous read-out or record of the measured in vivo signal, which is sometimes

difficult to trace by the physicist/QA personnel

•

bad positioning of the diode: not in the centre, too close to shielding blocks, etc. (cfr.

Section 2.1)

•

bad fixation of the diode

systematic

•

determine

calibration

factor

determine

correction

factors

convert

read-out

signal/dose

calculate exp.

read-out

signal/dose

record SSD

position

diode

accessories

in the beam

shielding

blocks

wedge

read-out

electrometer

•

individual

treatment unit:

treatment machine + table

•

R & V

simulator

•

simulator

TPS

•

TPS

prescription

•

treatment unit

set-up

patient

set-up

treatment preparation

treatment delivery

equipment breakdown

human errors

Figure 3.1 Schematic representation of error-prone steps in the quality control process

3.2.2 DEVIATIONS IN THE TREATMENT PROCESS (DOS IMETRIC ERRORS)

To facilitate the analysis of possible errors in the treatment chain, dosimetric errors are

divided into three categories:

•

human errors in data generation and data transfer

•

errors due to equipment breakdown or malfunctioning

•

positioning discrepancies between treatment planning and delivery

Figure 3.2 Schematic overview of the radiotherapy process.

3.2.2.1 ERRORS IN DATA GENERATION AND DATA TRANSFER (HUMAN ERRORS)

In Figure 3.2 a compact scheme is given of the radiotherapy process from prescription to

delivery. Each arrow in the diagram represents a transfer of data, which is error prone, and

every box may generate erroneous data. Depending on the organisation of the department

R & V
and/or

treatment chart

simulation

parameters

prescription

parameters

TPS

immobilisation

device

gantry angle

collimator

angle

field size

modality/

energy

MUs

(dose/fraction)

wedge

shielding

blocks

table

parameters

SSD

bolus

and on the possibilities of the available equipment, the practical information transferred

between the boxes may be different. At the end of the chain, before treatment delivery,

parameters are either recorded in a Record and Verify system (R & V) or written on the

treatment chart.

An erroneous transfer of prescribed dose from prescription to TPS can only be detected by

in vivo dosimetry if the entrance dose is calculated by hand or with an independent "home

made" system to predict the expected diode signal. If the planning target dose is - by

mistake - different from the prescribed target dose, the entrance dose calculated (manually)

with the prescribed dose will differ from the entrance dose calculated with the TPS, and

hence will be detected. The dark boxes in the last column of Figure 3.3 represent the other

parameters for

Figure 3.3 Overview of the parameters of the treatment preparation process

which errors can be detected by entrance in vivo dosimetry, whether or not the TPS is used

to calculate the entrance dose:

•

involuntary absence or presence of beam attenuators:

wedge

Presence or absence of the wedge in the beam can be considered as an important

cause of accidental treatment delivery (differences in absorption rates up to 60 %

for heavy wedges can occur). A discrepancy in the choice of the wedge (for

instance 45° instead of 30°) will also be detected by entrance in vivo dosimetry.

However, errors in the orientation of the wedge will not be detected by entrance in

vivo dosimetry (on the beam axis).

tray and blocks

Since the absorption rate of a tray holder is usually a few percent, its presence or

absence can be monitored. However, erroneous block positioning in the field is

unlikely to be detected by means of entrance in vivo dosimetry.

individual compensators

Errors in the positioning or in the choice of individual compensators may be

detected by in vivo entrance dose measurements.

•

treatment modality and energy

From a theoretical point of view, measuring the entrance dose is not a conclusive

method to detect errors in modality (photons, electrons) or in beam energy since

both the response of the diode and the dose rate of the various beams may be

different. However in practice these types of errors may also be detected as

reported for instance by Essers et al. [Essers 1999] for a dual energy linac.

Depending on the magnitude of the discrepancy between the prescribed field size and the

actual field size, the measured in vivo entrance dose theoretically shows an error in the

treatment delivery. In practice this error is so small that errors in field size are normally not

detected by in vivo dosimetry but only by portal imaging.

Before treatment execution, an independent check of data transfer (including MU

calculation) on the treatment chart and/or the R&V system should always be performed

[AAPM 1994]. It has been demonstrated that this simple tool significantly reduces the

incidence of human erro rs. However, even with this systematic check, in vivo dosimetry is

indispensable to trace a number of errors that otherwise would escape attention

([Calandrino 1997], [Duggan 1997], [Essers 1999], [Fiorino 2000]).

3.2.2.2 ERRORS DUE TO EQUIPMENT BREAKDOWN OR MALFUNCTIONING

As shown in Figure 3.2 radiotherapy departments may suffer from a breakdown or

malfunctioning of equipment at 4 levels which may translate into erroneous treatment

delivery: the simulator, the TPS, the R & V and the treatment unit (encompassing the

treatment unit itself, the treatment couch, fixation devices etc.). While a breakdown or

malfunctioning of the simulator is not so likely to be detected by in vivo dosimetry,

breakdown of the treatment unit and more specifically large variations in beam output can

easily be detected by in vivo entrance dose measurements. A typical other source of errors

which can be unveiled by entrance dose measurements is malfunctioning or incorrectly

used new software or a software upgrade for the calculation of MU ([Leunens 1993],

[Lanson 1999]). As far as the R & V system is concerned, malfunctioning is potentially very

dangerous if the system is also used for setting up the patient, and can be discovered by in

vivo dosimetry.

3.2.2.3 DISCREPANCIES IN PATIENT POSITIONING/GEOMETRY BETWEEN

TREATMENT PLANNING AND DELIVERY

Deviations in patient set-up at the time of treatment delivery can be due to random human

errors (especially if the treatment couch parameters are not verified by the R & V system

[Leunens 1993]) or to systematic machine-related errors (like a bad resetting of the “zero”

indicator of the simulator couch (Section 5.4 and [Fiorino 2000]), in which cases they

belong to the two previous categories of errors. Other sources of a wrong positioning,

however, are patient motion or a change in patient thickness due to swelling or shrinkage.

While entrance dose measurements give only direct information about the patient set -up

with respect to the beam (in particular an incorrect SSD), they also result in the detection of

patient thickness errors, if patient thickness is reassessed to trace the origin of a large

incorrect SSD.

3.3 EVALUATION OF IN VIVO DOSIMETRY DATA

3.3.1 ACTIONS AFTER THE FIRST MEASUREMENT

If the result of the first measurement is outside the tolerance/action level, a number of

controls should be activated (see Figure 2.2). This chain of controls involves checks in

order to reveal either quality control process errors (e.g. the immediate check of the diode

position, or the “a posteriori” recalculation of the expected signal) or real treatment process

errors (e.g. the immediate check of the patient position, or the data transfer control in the “a

posteriori” check).

First, an immediate check (i.e. with the patient on the treatment couch) of the treatment set -

up and treatment parameters must be performed. The most common errors are differences in

SSD and wrong positio ning of the diode, which can both be checked by the radiographers

on the spot. Differences in SSD due to the use of immobilisation devices or bolus should

be corrected by the appropriate inverse square law correction factor; in other cases the

SSD deviation signifies a real treatment process error (see Section 5.4). An evaluation of

the correct

positioning of the diode in the field centre can still be made afterwards on a

portal image (for treatment geometry verification), if this was taken simultaneously with the

diode measurement ([Essers 1994], [Lanson 1999]).

If no explanation of the discrepancy is found, further “a posteriori” checks should be

performed, possibly the same day as the in vivo control. The “a posteriori” check should

concern the data transfer control, checking the congruence of all technical and dosimetry

data of the treatment planning/simulator chart to the corresponding data on the treatment

chart (and/or R&V system). It must include the agreement between the effectively delivered

MU against the planned ones, the correctness of MU calculation and the correct use of

wedges and blocks. If the in vivo dose with blocked fields is lower than the expected one,

the diode position should be checked especially if the block is near the irradiation field

centre. In any case, it is a good practice to perform a second in vivo control check, also if

no apparent errors are found.

Tolerance levels generally coincide with action levels in most institutions. If t his is not the

case, performing merely a second in vivo dose measurement could be a limited, time -saving

action related to an out-of-tolerance deviation within the action level. A similar approach

can be followed if two different action levels are defined (“low” and “high”, for instance: 5

and 10 %). If the measurement is outside the “low” level, but still within the “high” action

level, the immediate check may for instance be avoided and just an “a posteriori” check

could be performed by the physicist.

3.3.2 PERSISTING DEVIATIONS: INTERPRETATION OF THE RESULT

A number of papers report that a lot of deviations exceeding the action level might not be

related to proven human errors or errors due to equipment breakdown ([Cozzi 1998],

[

Essers 1999

]

, [Heukelom 1991a], [Leunens 1990a], [Leunens 1991], [Loncol 1996], [Mangili

1999], [Millwater 1998], [Mitine 1991], [Noel 1995],

[

Voordeckers 1998

]

). On the other side a

second measurement reduces the probability of some quality control process errors such

as bad diode positioning. So, in the chain of control checks following in vivo dosimetry,

some deviations might not be attributed to a treatment process error nor to a quality

control process error. For such a situation the data should be discussed by the

physics/QA team and the origin should be traced.

In order to facilitate the search for the cause of the persistent deviation, an entrance dose

measurement with the diode and an ionisation chamber in a solid phantom in the same

clinical treatment conditions (SSD, field size , gantry and collimator rotation, block,

wedge…) may be useful. This is particularly true during the first phases of implementation

of systematic in vivo dosimetry procedures, when the accuracy of in vivo dosimetry in the

various clinical situations may not fully be assessed. If the deviation between the phantom

entrance dose measured with the diode and the calculated one is acceptable, the in vivo

deviation could be attributed to a difficulty in the diode positioning on the patient’s skin. If

the dose meas ured with the ionisation chamber is in agreement with the expected one, but

the diode reading is not correct, the deviation of the diode signal could be explained by a

bad calibration or wrong correction factors. If the phantom entrance dose measured with

the ionisation chamber is not in agreement with the expected dose, dose calculation

mistakes might be present.

In general, the main causes of persistent deviations are:

•

difficulties in setting up the patient: these are more likely to be detected if diode

readings are not corrected by inverse square correction factors

•

uncertainty in diode reading due to critical positioning (tangential beams, wedged

beams,…) or to lack of correction factors

•

bad electrometer/diode calibration

•

erroneous calculation of the entrance dose by the TPS

When the cause of the discrepancy is identified, an action may be required for the single

patient such as checking patient positioning at the simulator and/or checking the patient’s

thickness. If persistent deviations sytematically occur for a certain configuration of beams

(for instance wedged fields), more accurate assessment of diode correction factors or

further investigations on the accuracy of dose calculation may be required. After reviewing

a large set of data, a high rate of second checks/persistent deviations for a certain

configuration of fields may also suggest a modification of the tolerance/action level for

such a category.

3.3.3 MONITORING IN VIVO DOSIMETRY WITH TIME

After implementing a procedure for systematic in vivo dosimetry, it is very important to

continuously monitor the adequacy and the efficacy of the system. A periodic review of

the database of in vivo dosimetry data with some statistical analysis is very useful to drive

the physicist and the clinician in adjusting the procedures to the real local conditions. An

important goal is the verification of the adequacy of tolerance/action levels: a too high rate

of second checks may have a negative impact on the operators and efforts should be made

to reduce the additional wo rkload, while maintaining an acceptable action level.

Continuous monitoring of in vivo dosimetry data may therefore indicate the need for

adjustment with time of tolerance/action levels.

Statistical analysis of the deviations between expected and measured entrance dose may

give information suggesting possible fields of investigation and/or possible improvements

of the quality control process (for instance, more accurate assessment of diode correction

factors, new schedule for diode calibration etc.).

Although some errors or inaccuracies may also be detected on an individual basis, they

will be clearer with a statistical approach because of the existence of fluctuations in the

measuring procedures. Subgroups of patients can be pooled for instance breast, hea d &

neck, brain, etc. The distribution of the deviations has been shown to reveal systematic

errors linked to specific treatment techniques or to calculation methods. Some relevant

examples concerning large cohorts of patients are given in the literature [Noel 1995],

[Leunens 1990a], [Leunens 1990b], [Leunens 1991] and [Fiorino 2000] and in the single

institution’s experiences as reported in Chapter 5. It should be realized that this very

CHAPTER 4 TECHNIQUES AND PROCEDURES IN DIFFERENT

RADIOTHERAPY CENTRES

The following information concerning clinical in vivo dosimetry procedures (not

exclusively oriented towards entrance dose measurements) is the result from a

questionnaire that has been sent to centres having experience with routine in vivo dose

measurements. Apart from the institutions which have co-operated for this booklet, the

‘Nederlands Kanker Instituut’ from Amsterdam and the ‘Centre Alexis Vautrin’ from Nancy

have provided information (see also [Essers 1993, 1994, 1995a and 1995b], [Heukelom

1991a, 1991b, 1992, 1994], [Lanson 1999] for Amsterdam and [Noel 1995] for Nancy).

4.1 WHAT EQUIPMENT DO YO U USE TO CARRY OUT ROUTINE IN VIVO DOSE

MEASUREMENTS?

LEUVEN

BARCELONA

Irradiation equipment

- X-rays : 6, 18 MV

(Saturne 40, 42, GE, Clinac 2100, Varian)

Irradiation equipment

- X-rays 6, 18 MV / e

6 - 16 MeV

(Clinac 1800, Varian)

In vivo dosimetry equipment

DPD6, DPD3, DPD510 (TBI) electrometer

(Scanditronix)

- Diodes (Scanditronix)

6 MV

EDP-20, EDP-20+1 mm Cu

18 MV

EDP-20, EDP-20+1 mm Cu

TBI

EDE

In vivo dosimetry equipment

- DPD510 electrometer (Scanditronix)

- Diodes

6 MV

EDP-10 (Scanditronix)

18 MV

EDP-30 (Scanditronix)

QED 1116 (Sun Nuclear)

P30 (Precitron)

Isorad-p 1164 (Sun Nuclear)

electrons

EDD-2 (Scanditronix)

TBI

EDP-30 (Scanditronix)

NANCY

COPENHAGEN

Irradiation equipment

Co (Th780, AECL)

- new multimod. LINAC

(X-rays 6, 10, 25 MV, Saturne 43, GE-CGR),

(X-rays 6, 10 MV, SL15, Elekta),

(X-rays 6, 25 MV, Clinac 23EX, Varian)

Irradiation equipment

X-rays 4, 6, 8 and 18 MV

(Varian Clinac 600C, Clinac 2100C, Clinac

2300CD)

In vivo dosimetry equipment

- DPD3, DPD5 or DPD6 electrome ter

detection and diodes assembly

(Scanditronix)

- Diodes

p-type diodes except for two n -type

diodes with additional correction factors:

cobalt

6 MV

HE or -10 (Scanditronix)

10 MV

EDP-10 (Scanditronix)

25 MV

EDP-20 (Scanditronix)

(electronic equilibrium if necessary

obtained with bolus)

- Apollo5, Apollo10 electrometer detection

and diodes assembly (Precitron AB) for

TBI

In vivo dosimetry equipment

- electrometers:

Apollo 5 (Precitron)

- Diodes

4 MV

P10 (Precitron),

6 MV

P10 (Precitron),

QED 1115 (Sun Nuclear)

Isorad-p 1163 (Sun Nuclear)

8 MV

P20 (Precitron)

Isorad-p 1163 (Sun Nuclear)

18 MV

P30 (Precitron)

QED 1116 (Sun Nuclear)

AMSTERDAM

MILANO

Irradiation equipment

in vivo dose measurements performed for

- X-rays 6 and 8 MV (Philips SL15/SL25)

- X-rays 4 MV (ABB)

Irradiation equipment

- X-rays 6 MV (Linac 6/100)

- X-rays 18 MV (Linac 1800, in vivo

dosimetry to be implemented)

In vivo dosimetry equipment

- p-type diodes (Scanditronix EDP-20)

coupled to a custom-built diode measuring

system (hardware (electrometer) and

software (diode measurement files)

In vivo dosimetry equipment

- DPD510, DPD3 electrometer

(Scanditronix)

- Diodes

EDP-10, EDP-30 (Scanditronix)

EDINBURGH

Irradiation equipment

X-rays 6, 8, 9, 15, 16 MV / e

5-20 MeV

(Varian 600, 600CD, ABB CH6, CH20, RDL Dynaray 10)

In vivo dosimetry equipment

DPD6, DPD3, DPD510 electrometers (Scanditronix)

Diodes (Scanditronix)

6 MV:

EDP-10, EDP-10 + 0.6 mm brass

8,9 MV:

EDP-20

15, 16 MV:

EDP-20, EDP-20 + 1.2 mm brass, copper

electrons

EDE, EDD-5, EDD-2

Mounting: home -built quick-swing ceiling mounted system; being rolled out to all

machines as new machines being installed

4.2 PHILOSOPHY OF YOUR DEPARTMENT CONCERNING THE USE OF IN VIVO

DOSIMETRY?

4.2.1 WHEN DO YOU USE IN VIVO DOSIMETRY?

LEUVEN

BARCELONA

- for every patient at first treatment session

(simultaneous with portal film)

during treatment when major changes in

irradiation parameters take place (after a

new monitor unit calculation)

- TBI treatments: first session

- for every patient treatment at the second

treatment session, and when there is a

treatment modification (first fraction:

portal film)

- TBI treatments: all sessions

NANCY

COPENHAGEN

- for every patient at the second or third

session (first fraction: portal film)

- during treatment whenever major changes

in irradiation parameters take place

(reduction of field size, blocks, wedge)

- TBI treatments: all sessions

- intention of including every patient

(achieved to 90%, still in the

implementation phase)

- within the third treatment session

AMSTERDAM

MILANO

- for two special treatment techniques with

high dose/high precision protocols:

parotid gland and prostate irradiation;

measurements are performed during two

different treatment sessio ns

- in the near future, all treatment techniques

will be checked systematically one by one

- for every patient, within the first week of

treatment. during treatment whenever

major changes in irradiation parameters

take place (reduction of field size, blo cks,

wedge)

- TBI treatments: first treatment

EDINBURGH

- all treatment machines and treatment techniques are systematically checked in detail on

sufficient patient numbers to give a statistically valid study. From this systematic

deviations are identified and corrected, random deviations are quantified; decisions are

then taken on how to implement in routine use

- in routine use, typically for every patient within the first week, or after significant

changes in treatment. Currently not on all machines; this is being rolled out to all

machines as new machines are installed

- plus, infrequent audits on a selection of patients for a given machine, site and

technique, which repeat the initial pilot studies on a small group of patients

- electron measurements ju st beginning

4.2.2 WHAT DO YOU MEASURE?

LEUVEN

BARCELONA

- entrance dose

- for opposed photon beams: target

absorbed dose by measurement of

entrance and exit doses. For each beam

the target dose is calculated from midplane

dose using a ratio of PDD. The midplane

dose is calculated as the mean multiplied

by an experimental correction factor or by

a Rizzotti approach [Rizzotti 1985]

- for non-opposed photon beams: entrance

dose

- for electrons: entrance dose measurements

NANCY

COPENHAGEN

- target absorbed dose by measurement of

entrance and exit doses

- usually midline dose is calculated; else the

midline dose is for each beam converted to

the dose at the specified point using

PDDs before summing the contributions

of the different beams

- entrance dose

AMSTERDAM

MILANO

- target absorbed dose by measurement of

entrance and exit doses

- for prostate treatments, the target

absorbed dose is converted to midline

dose by using a modified Rizzotti

approach [Rizzotti 1985] and then the dose

in the prescription point is calculated by a

simple PDD algorithm. The contributions

of the different beams are summed

- for parotid gland: target absorbed dose by

measurement of entrance dose plus PDD

correction

- entrance dose

- TBI: midline dose distribution (chest,

abdomen, pelv is) by combining diodes

and portal films (transit dosimetry) data

4.3 PROCEDURE FOR IN VIVO DOSIMETRY?

4.3.1 CALIBRATION PROCEDURE? WHICH CORRECTION FACTORS ARE USED?

LEUVEN

BARCELONA

Equipment

- polystyrene phantom

Calibration

- ref. conditions:

SSD 100 cm,

field size 10 x 10 cm

- for absolute dose determination with the

ionisation chamber, the Dutch (NCS)

protocol is used, however, without

displacement factor (see Section 1.2)

- TBI: calibration factors in TBI conditions

Correction for

- temperature (only for TBI)

- SSD, field size, wedge, tray

Equipment

- polystyrene phantom

- Plastic Water

phantom (CIRS)

- water phantom equipped with thermostat

Calibration

- ref. conditions: SSD 100 cm, 22.5°C

field size 10 x 10 cm

- for absolute dose determination with the

ionisation chamber, the Spanish protocol

is used, however, without displacement

factor (for calibration of entrance dose

measurements) (see Section 5.1.2)

- TBI: entrance and exit calibration factors in

TBI conditions

Correction for

- temperature (only for TBI)

- entrance: field size, SSD, tray, wedge,

directional dependence

- exit: none (< 1%)

NANCY

COPENHAGEN

Equipment

- polystyrene phantom

Calibration

- ref. conditions:

SSD 100 cm,

SSD

Co 80 cm

field size 10 x 10 cm

- TBI: entrance calibration factors in TBI

conditions

Correction for

- variation of response of exit detector with

dose rate

- wedge correction fa ctor

Equipment

- Solid Water

phantom ( RMI 457)

Calibration

- initially against NE Farmer chamber

- periodic calibrations: against the Clinac

monitor chamber in connection with

output check of the treatment machine

Correction for

- field size, SSD, tray, layers of

compensation filter, temperature and

directional dependence (no wedge

correction because of dynamic wedges)

AMSTERDAM

MILANO

Equipment

- polystyrene phantom

Calibration

- ref. conditions: SSD 100 cm

field size 15 x 15 cm

15 cm thick phantom

- Dutch (NCS) protocol without

displacement factor

Correction for

- patient thickness, SSD, field size, wedge +

shift of measurement point in wedge

direction [Essers 1994], angle, air gap,

Equipment

- acrylic phantom

Calibration

- ref. conditions: SSD 100 cm,

field size 10 x 10 cm

- TBI: treatment conditions

Correction for

- SSD, field size, wedge, tray,

directional dependence, (temperature)

- TBI: off-axis

carbon fibre table, temperature

EDINBURGH

Equipment

- Solid Water

phantom (RMI 457)

Calibration

- ref. conditions, 100 SSD, 15 x 15 cm

field, 15 cm thick phantom

- absolute dose determination using calibrated ionisation chamber using UK protocol:

for entrance dose calibration, take out the displacement correction

for exit dose calib ration, add in an average ‘build -down’ correction

Correction for

- measured for every parameter: e.g. SSD, field size, phantom/patient thickness, directional

dependence, temperature, wedge, tray, (for both entrance and exit initially)

- build-up caps used on diodes to minimise the range of values of correction factors

- detailed correction factors used for initial systematic studies and for audits

- for routine use, mid -range correction factors for the irradiation parameters used for a

specific technique and treatment machine are combined into ‘generic’ correction factors

to be applied for that particular treatment and machine

4.3.2 WHICH MEASURED AND EXPECTED DOSES ARE COMPARED?

LEUVEN

BARCELONA

- the expected entrance dose is the dose

calculated by the TPS

- the expected doses (entrance, exit and

ICRU point) are the doses calculated by

the TPS;

- for electrons, the expected dose is the

dose calculated at dose maximum (which is

also the prescribed dose)

NANCY

COPENHAGEN

- the expected dose is the dose calculated at

the ICRU dose specification point, or the

prescribed dose, if there is no isodose

distribution available

- the expected entrance dose is the dose

calculated either with an independent

spread sheet program or by the TPS

AMSTERDAM

MILANO

- the exp ected dose is the target absorbed

dose, calculated with the TPS

- the expected entrance dose is calculated

by an independent formula

(implemented on PC)

EDINBURGH

- the expected entrance dose is either that from the TPS, or calculated manually,

dependin g on treatment

- the expected exit dose is from the TPS, or calculated manually depending on treatment

- the expected target volume dose is that calculated at the specification point, or the

prescribed dose if there is no isodose distribution (including electrons)

- for breast the expected dose is taken from the plan at 1.5 cm below the diode

measurement point

- for routine use, expected diode reading ranges (expected reading is expected dose

divided by calibration factor and by generic correction factor) are supplied to the

treatment machine by physics/planning, so that the radiographers only have to tick a

box that the reading is within range or not

4.3.3 VALUE OF TOLERANCE AND ACTION LEVELS + ACTIONS UNDERTAKEN

LEUVEN

BARCELONA

Tolerance level:

5 % for 6 MV

10 % for 18 MV

1st action level:

5 % for 6 MV

10 % for 18 MV

2nd action level:

10 % for 6 MV

15 % for 18 MV

Actions

Cfr. Flow chart of Figure 2.2

Tolerance level: 5 %

Action level: 5 %

Actions

immediate action:

∆ ≥ 5 %

- radiographer

- immediate check of treatment parameters

(MU, field geometry, patient position,

movement of diode)

a posteriori action

∆ ≥ 5 %

- check of all parameters

- IVD at the next treatment session

- if

∆

≥

5 % persists: simulation of treatment

with Plastic Water

phantom, with diode

and ionisation chamber at the same time

NANCY

COPENHAGEN

Tolerance level:

5 %

1st action level: 5 %

2nd action level: 10 %

Actions

Immediate action:

∆ (entrance dose) ≥ 10 %

- radiographer (+ physicist)

- immediate check of treatment parameters

a posteriori action (physicist):

∆ (target dose) ≥ 5 %

∆

≤

10 %

- verification of the MU calculation

- thorough investigation of all treatment

parameters

request for IVD at the next treatment

session

∆

≥

10 %

request for IVD at the next treatment

session in presence of physicist

Tolerance level: 5 % or 8% depending on

complexity of treatment

Tolerance and action levels coincide.

Actions

- check of quality control pro cess

(calculation of the expected dose,

positioning of the diode, practical

problems)

- check of treatment preparation process

(dose calculation and treatment chart data)

- repeat diode measurement

- if deviation persists: simulation of

treatment with Solid Water

phantom,

with diode and ionisation chamber at the

same time.

AMSTERDAM

MILANO

Tolerance level:

- 2.5 % for the prostate

- 4.0 % for the parotid gland

Action level:

- 2.5 % for the prostate

- 4.0 % for the parotid gland

Actions

- immediate check of treatment parameters

- a third measurement is performed if only

one measurement, and the average, is

exceedingthe action level

- if deviation persists: patient dose (MUs) is

always corrected for the other treatment

fractions and the origin is traced by

additional phantom measurements and

calculations with the TPS

Tolerance level:

- 5 %

- 7 % for tangential wedged fields

Action level:

- 5 %

- 7 % for tangential wedged fields

(see Section 5.5)

Actions

immediate check (technician):

- always an independent check of treatment

parameters (including SSDs): the operator

performing the check is different from the

radiographer who sets up the patient

a posteriori action (physicist):

∆ ≥ 5 %

(7 % for tangential wedged fields)

- check of all parameters (data transfer, dose

and MU calculation)

- IVD at the next treatment session

∆

≥

5 % persists, possible measurement

in acrylic phantom with ionisation

chamber (and diode)

4.3.4 TIME PERIOD BETWEEN CHECKS OF CALIBRATION AND CORRECTION

FACTORS

LEUVEN

BARCELONA

Calibration

- every month

Correction factors

- once a year

Calibration

- every 50 Gy of accumulated dose

- TBI: every four TBIs

- electrons: in evaluation

Correction factors

- temperature: every 6 months

- FS, tray, wedge, angle: every year

NANCY

COPENHAGEN

Calibration

- once a week

- TBI: before every first session

Calibration

- every third month

Correction factors (in evaluation)

AMSTERDAM

MILANO

Calibration

- every two weeks, depending on the

amount of accumulated dose

Correction factors

- twice a year

Calibration

- every month

Correction factors

- once a year

EDINBURGH

Calibration

test calibration quarterly; but beginning to use diodes as routine daily treatment machine

dose consistency check. In this case, checked versus ion chamber weekly

Correction factors

annually, or if unexpected changes in calibration

4.4 WHAT SYSTEM DO YOU USE FOR THE RECORDING OF IN VIVO DOSE

MEASUREMENTS?

LEUVEN

BARCELONA

- manual recording of the diode signal on a

separate in vivo sheet

- manual recording of the measurements on

a separate in vivo sheet (not included in

patient record) with relevant information:

- field size, wedge, tray, SSD

- ICRU point depth

- PDD at ICRU point depth

- PDD at entrance

- PDD midplane

- PDD at exit

for pelvic treatments and TBIs, the

measurements are entered in a “Excel

Book” containing the correction factors,

with the irradiation parameters and the

expected doses;

the corrected doses and the deviations

between expected and measured entrance,

exit and prescribed doses are calculated

automatically; statistical analysis of the

data is performed automatically

NANCY

COPENHAGEN

- manual recording of the measurements in

the treatment chart (relevant information

entered in a file by the physicist):

localisation

treatment unit

- manual recording of the measurements

and relevant beam data in a database for

statistical evaluation

treatment technique

beam geometry

wedge

immobilisation technique

ratio meas./calc. entrance dose for

each field

ratio meas./calc. target absorbed

dose for whole treatment session)

future implementation of recording in record

and verify system

TBI: measurements are recorded in real time

from the electrometer and measured

absorbed doses at 7 points of interest

are computed by home -made PC

software

AMSTERDAM

MILANO

- the diode measurement system prepares a

diode measurement file for every patient

field, containing beam, patient and diode

parameters (for instance calibration and

correction factors are determined for each

- manual recording of the in vivo dosimetry

results with a number of relevant

information (see Section 5.5)

- periodic update of Excel files for statistical

analysis

EDINBURGH

- for systematic measurements and for full audits, manual recording of results on

separate in vivo sheet

- for routine use, expected reading range is recorded on the treatment sheet (in a short

in vivo section) and the radiographers check a tick box

If not within the expected range, the sheet is drawn to the attention of the physics

group

4.5 WORKLOAD? SPECIFIC TASKS OF PEOPLE INVOLVED?

LEUVEN

- implementation/ maintenance related tasks

- initial acceptance/calibration/

QA person in charge

10 hours/diode

correction factors

- periodic calibration

QA person in charge

45 min/diode

- periodic determination

QA person in charge

4 hours/diode

of correction factors

- training of radiographers

QA person in charge

- patient related tasks

- calculation of expected

dosimetrist

3 min/patient

diode signal

- measurements

radiographer

3 min/patient

- out-of-tolerance analysis

QA person in charge

1 hour/week

+ physicist

for 2000 patients/year

- TBI (single)

physicist

1 hour

- TBI (fractionated)

physicist

5 min/session

BARCELONA

- implementation/ maintenance related tasks

- initial acceptance/calibration/

physicist

4 hours/diode

correction factors

- periodic calibrations

physicist

30 min/2 weeks

- training of radiographers

physicist

and dosimetrists

- patient related tasks

- recording in in vivo sheet for

dosimetrist

10 min/patient

every field FS, ICRU point depth,

PDD at ICRU point, at entrance,

at midplane, and at exit

- measurements

radiographer

3 min/patient

- recording of data in Excel book

physicist

5 min if there are

and daily evaluation

no problems

- TBI (hyperfractionated)

physicist

45 min during 3 days

NANCY

- patient related tasks

- measurements

radiographer

3 min/patient

- evaluation

physicist

1 hour/day

- TBI

physicist

10 min/session

COPENHAGEN

- implementation/ maintenance related tasks

initial calibration procedures

QA physicist

5 hours/diode

periodic calibration

QA physicist

10 min/diode

training of involved personnel QA physicist

30 min/month

- patient related tasks

calculation of expected signal dosimetrist

5 min/patient

measurements

radiographer

3 min/patient

- out-of-tolerance analysis,

physicist

average 5 min/patient

documentation, evaluation

AMSTERDAM

- total workload

0.4 full time equivalent (2 days work/week)

- time per patient

4.4 hours

- making appointments

12 min

- measuring (twice) + waiting;

120 min

the measurement itself takes up to

5 min

- analysis

30 min

- additional measurements

18 min

- storing of data

30 min

- implementation of corrected MU

6 min

- analysis of discrepancies

48 min

- additional time per week

- phantom tests

2 hours

- administration

1 hour

- consultation with co-workers

2 hours

- consultation with co-workers

2 hours

MILANO

- implementation/ maintenance related tasks

initial calibration procedures

physicist + t echnician

4-6 hours/diode

periodic calibration

physicist + technician

10-15 min/month/diode

- patient related tasks

patient data collection

technician

5-10 min/patient

measurements

technician

5-15 min/patient

entrance dose calculation,

physicist

average 5 min/patient

comparing to measured dose

- data analysis

physicist + technician

up to 1 hour/patient

(including transit dosimetry,

routinely for TBI patients)

- TBI measurement

technician

1 hour/patient

EDINBURGH

- implement ation/ maintenance related tasks

- initial acceptance/calibration/ physicist/physics technician

5 hours/diode

correction factors

- periodic calibrations

physicist/physics technician

1 hour/quarter

year/diode

- patient related tasks (routine use)

- calculate expected signal

physics/planning personnel a few minutes

measurements

radiographer

less than 5 min/patient*

- Co-ordination currently by part -time research radiographer/dosimetrist seconded to

the

physics group.

the

physics group.

-* methodology designed so that this is in parallel with other tasks, i.e. adds minimal

time

to patient treatment

- systematic studies and audits

Physicists and seconded trainees/project students (including physics, radiographers,

radiation oncologists). Significant times involved.

4.6 EXAMPLES OF PRACTICAL PROBLEMS?

LEUVEN

BARCELONA

- wrong entrance dose calculation by TPS

- some important errors are not traced by in

vivo dose measurements (for instance

errors in TPS target dose, since the

expected entrance dose is the one

calculated with the TPS)

- use of a polystyrene phantom (for

calibration, determination of correction

factors, determination of Rizzotti curves

and of correction factors for midplane

calculation) instead of a water phantom

- use of cerrobend (block) correction factor

- exit correction factors: how can they be

measured to guarantee independence of

factors and independence of the phantom

thickness?

- non-symmetric heterogeneities

NANCY

COPENHAGEN

- positioning of diodes:

- chest wall irradiation

- in presence of immobilisation

device, use of support (head support, ...)

- positioning of diodes:

- presence of immobilisation

- near blocks

- half-beam technique with wedges

- near blocks

- small fields (exit diode)

- estimation of midline dose:

presence of wedge

presence of heterogeneities

isocenter not situated at mid -depth

non-opposed beams

different X-ray energy used for the

same patient (ant 6 MV/post10

MV)

- mounting of equipment:

convenient handling of the diodes and the

cables

- calibration:

difficulties with narrow sensitivity range

of electrometers => improper matching to

the diode sensitivity (during

implementation)

AMSTERDAM

MILANO

all encountered problems turned out to be

real problems with the treatment planning

system or the performance of the linear

accelerator

action level for tangential wedged beams

modified with time after statistical analysis

EDINBURGH

- positioning on some surfaces, particularly on steep angles

- measurement problems when the beam is incident through the couch or head support,

etc.

- what temperature correction should be applied in certain situations

- measurements below bolus

- measurements close to blocks (particularly CRT blocks), asymmetric fields, etc.

- limitation of resolution of electrometer in small reading situations, particularly for small

wedged components of fields on motorised wedge machines

- handling of diodes and cables in rooms where our ceiling mounted support is not yet

installed: particularly connector failures

CHAPTER 5

EXPERIENCES FROM DIFFERENT RADIOTHERAPY

CENTRES

In order to provide more detailed examples regarding the implementation and the

functioning of in vivo dosimetry in clinical routine, we have selected contributions from the

authors’ institutions about an aspect of in vivo dosimetry that they have worked on

specifically. Some contributions offer reference data concerning basic procedures, from

diode calibration to evaluation of the measurements; others contain specific suggestions

for improvement or refinement of procedures. More details and data can be found in [Jornet

2000], [Lööf 2001], [Georg 1999] and [Fiorino 2000].

5.1 CALIBRATION AND MEASUREMENT PROCEDURES – THE BARCELONA

EXPERIENCE

This section lists detailed examples of methods and results of the calibra tion procedures,

as explained in Section 1.2, performed in the radiotherapy department of the Hospital Santa

Creu I Sant Pau in Barcelona. A recent study [Jornet 2000] concentrates on the

performance of p-type and n-type diodes in high energy beams, which will be elucidated in

some detail in this overview (see also Section 5.2).

5.1.1 TESTS PERFORMED BEFORE DIODE CALIBRATION

Due to the way diodes are made, two diodes even from the same fabrication batch may

behave differently when irradiated. Therefore it is recommended to perform some tests

before using them in routine. The results of these tests should be compared with the

technical specifications provided by the manufacturers.

The tests performed whenever a new diode is received in our centre are:

1. Signal stability after irradiation

2. Intrinsic precision

3. Study of the response/dose linearity

4. Verification of the equivalent water depth of the measuring point (water

equivalent thickness of the build -up cap)

5. Perturbation of the radiation field behind the diode

Some tests (4 and 5) are only performed for the first 3 or 4 diodes of a particular type. All

diodes are connected to the same channel of the electrometer to avoid drifts and loss of

signal, which depends on the channel to which they are connected. All channels of the

electrometer are checked regularly. The measurements corresponding to these tests are

performed at reference conditions (i.e. collimator opening 10 cm x 10 cm, phantom surface at

the isocentre). For most of the tests the diode is fixed on the surface of a plastic phantom,

i.e. a Plastic Water

phantom (CIRS).

The results of the tests for different diodes are summarised in Table 5.1.

EDP-10

(6 MV)

EDP-30

(18 MV)

P30

(18 MV)

QED

(18MV)

Isorad-p

(18MV)

1 stability after

irradiation (5 min)

0.3 %

-0.58 %

0.33 %

-0.06 %

-0.20 %

2. intrinsic precision

(SD)

(10 irradiations)

0.06 %

0.16 %

0.05 %

0.07 %

0.10 %

3. linearity

response/dose (r

)

1.0000

(0.2 - 7 Gy)

1.0000

(0.2 – 7 Gy)

1.0000

(0.2 - 3.5 Gy)

1.0000

(0.2 - 7 Gy)

1.0000

(0.2 – 7 Gy)

4. depth of diode

measuring point

(water equivalent

depth)

0.80 cm

1.4 cm

3.0 cm

2.2 cm

3.3 cm

5. dose decrease at 5

cm depth

6 %

3 %

9 %

6 %

14%

designed with a cylindrical build -up cap

Table 5.1 Overview of the results of the initial tests for different types of diodes

5.1.1.1 SIGNAL STABILITY AFTER IRRADIATION

The signal immediately aft er irradiation is compared to the signal five minutes after the end

of the irradiation. Five minutes is considered as the average of the time periods

encountered in clinical practice.

5.1.1.2 INTRINSIC PRECISION

The standard deviation of 10 readings of 100 MU each is calculated.

5.1.1.3 STUDY OF THE RESPONSE/DOSE LINEARITY

We verify that the response is linearly proportional to the absorbed dose for clinical

significant doses. As we verify the linearity between MU and dose regularly, we verify the

linearity of the system diode-electrometer between 15 and 600 MU.

5.1.1.4 VERIFICATION OF THE WATER EQUIVALENT DEPTH OF THE MEASURING

POINT

The diode is fixed on the surface of the Plastic Water

phantom and covered with a

special Plastic Water

slab to avoid air gaps (Figure 5.1). Irradiations with X-rays are

performed while adding Plastic Water

slabs on top of the diode until the reading reaches

a maximum. As the depth of dose maximum in water at these irradiation conditions is

known, the water equivalent thickness of the build -up cap can be calculated (Figure 5.2).

Figure 5.1 Experimental set-up for the

determination of the water equivalent

thickness of the build-up cap

Figure 5.2 Diode signal at 18 MV as a

function of the thickness of the Plastic

Water

slabs on top of an EDP-30 diode.

The depth of dose maximum for a 10 x 10

field and 18 MV X-rays is 3.5 cm, so

the water equivalent thickness of the

build-up cap of EDP30 is 1.4 cm

5.1.1.5 PERTURBATION OF THE RADIATION FIELD BEHIND THE DIODE

One X-Omat V Kodak film is placed inside a Plastic Water

phantom at 5 cm depth. The

diode is fixed on the surface of the phantom and an irradiation is performed. Another film is

exposed under the same conditions, but without the diode. The dose decrease at 5 cm

depth underneath the diode is calculated by comparing the two beam profiles at this depth.

We use a film scanner (Scanditronix, an option of our field analyser RFA -300) to obtain

beam profiles. It has a spatial resolution of 0.1 mm.

5.1.2 DIODE CALIBRATION (ENTRANCE DOSE)

Diodes are calibrated against an ionisation chamber placed at the depth of dose maximum

inside a plastic phantom (polystyrene or Plastic Water

). The cylindrical ionisation

1.46

1.48

1.5

1.52

1.54

1.56

1.58

1.6

1.62

0.5

1.5

2.5

plastic water thickness on top of diode (cm)

diode reading

SSD = 100 cm

chamber (0.6 cm

) (IC) has a calibration factor traceable to the National Standard Dosimetry

Laboratory in Spain. The diodes are taped on the surface of the phantom near the field

centre, in such a way that they do not perturb the response of the ionisation chamber.

The calibration is performed in reference irradiation conditions (field size at the isocentre 10

x 10 cm

, SSD = 100 cm) (see Section 1.2). As the accelerator rooms are equipped with air-

conditioner, the room temperature is always between 21ºC and 22ºC. First, the reading -in-

plastic is converted to a reading-in-water by multiplying the reading-in-plastic with an

experimentally determined factor (in the case of Plastic Water

, this factor is 1). To

determine absolute dose-to-water, the Spanish dosimetry protocol is used. This includes

the application of a displacement factor for entrance dose. As the measurements are not

performed on the exponential part of the curve but at the depth of dose maximum, this

factor is not applied. The calibration factor F

cal

is determined as the ratio of the absorbed

dose determined with the ionisation chamber and the diode reading (see Section 1.2.2).

As the sensitivity of the diodes depends on dose rate, energy and temperature, some

correction factors will have to be applied to the diode reading when measuring conditions

differ from calibration conditions. Some of the correction factors depend, in addition, on

the diode calibration methodology. Since the diode is fixed on the patient’s skin, the scatter

conditions seen by the diode are obviously not the same as the scatter conditions seen by

the ionisation chamber. Therefore a field correction factor, for example, will have to be

applied even if the diode build -up cap is thick enough to guarantee electronic equilibrium.

Furthermore, as the dose rate sensitivity dependence may change with accumulated dose,

the correction factors that account for this dependence, such as the SSD correction factor,

will have to be checked regularly.

We determine the following correction factors:

1. Field size correction factor (CF

field size

)

2. Tray correction factor (CF

tray

)

3. Wedge correction factor (CF

wedge

)

4. SSD correction factor (CF

SSD

)

5. Angle correction factor (CF

angle

)

6. Temperature correction factor (CF

temperature

)

In addition, for some types of diodes, we performed tests to assess the importance of the

following issues:

7. Influence of the dose rate on the diode’s sensitivity

8. Sensitivity variation with accumulated dose

The results of the measurements of the entrance correction factors are given in Table 5.4

and Table 5.5

5.1.2.1 FIELD SIZE CORRECTION FACTOR (CF

)

The field size correction factor is defined as:

)

(

)

(

diode

(5)

where OF is:

)

(

)

(

)

(

(6)

with c the side of the square field in cm, and R the reading.

If the measurements of OF

diode

are performed at the same time as the measurements of OF

using a plastic phantom, attention should be paid to OF

because it may differ from OF

measured in water, so a factor to convert reading-in-plastic to reading-in-water should be

applied. This factor will probably depend on field size. To simplify things, one can measure

diode

and compare it with OF

measured in water at the depth of dose maximum.

Field size correction factors obtained for different diodes in different beam qualities are

shown in Figure 5.3 and Figure 5.4.

Figure 5.3 CF

for EDP-10 diodes and 6 MV X-rays. The mean and SD of measurements

performed with ten diodes are given.

Figure 5.4 CF

for EDP-30, P30, QED and Isorad-p diodes in an 18 MV X-ray beam. The

mean and SD for three diodes of each type and three measurements are shown.

0.980

0.985

0.990

0.995

1.000

1.005

side of square field (cm)

field size

EDP10

0.950

0.970

0.990

1.010

1.030

1.050

side of square field (cm)

field size

EDP30

P30

QED

Isorad-p

5.1.2.2 TRAY CORRECTION FACTOR (CF

tray

)

Shielding blocks are positioned on a tray attached to the treatment head. In our hospital,

the tray for the Clinac accelerator is made of PMMA of 0.5 cm thickness. Inserting a tray

between the source and the patient changes the amount of electrons that reaches the

patient’s skin. Therefore, if the diode does not have an appropriate build -up cap, the tray

correction factor varies with field size.

To determine CF

tray

, we measure the tray transmission for different field sizes at the depth

of dose maximum, first with an ionisation chamber and then with the diodes taped to the

surface of the plastic phantom. The transmission factors measured with the ionisation

chamber and with the diodes are compared, and CF

tray

as a function of field size is obtained

(Figures 5.5 and 5.6).

)

(

transmissi

)

(

transmissi

diode

tray

(7)

Where the transmission is defined as:

)

(

)

tray

(

)

(

transmissi

(8)

with c the side of the square field in cm, and R the reading.

5.1.2.3 WEDGE CORRECTION FACTOR (CF

wedge

)

Inserting a wedge in the beam results in a decrease of the dose rate and a hardening of the

spectrum of the beam. Therefore, as the sensitivity of the diode depends on both dose rate

and energy, a correction factor different from 1 is expected when using wedges.

The wedge correction factor is defined as the ratio between the wedge transmission factor

for a 10 x 10 cm

field, measured with the ionisation chamber placed at the depth of dose

maximum, and the wedge transmission factor for the same field size, meas ured with the

diode placed at the field centre taped on the surface of the phantom.

Figure 5.5 CF

tray

for EDP-10 diodes and 6 MV X-rays. The mean and SD of CF

tray

determined for 10 EDP-10 diodes are given.

Figure 5.6 CF

tray

for EDP-30, P30, QED and Isorad-p diodes in an 18 MV X-ray beam.

The mean and SD for three diodes of each type and three measurements are

shown.

0.975

0.98

0.985

0.99

0.995

1.005

1.01

side of the square field (cm)

tray

EDP10

0.970

0.980

0.990

1.000

1.010

side of square field (cm)

tray

EDP30

P30

QED

isorad-p

diode

max

wedge

)

10x10

(

transmissi

)

10x10

(

transmissi

(9)

with w the wedge angle.

For 6 MV X-rays CF

wedge

was determined for 10 EDP-10 diodes, three times each and for

different field sizes. The estimated uncertainties associated with the determination of this

factor are up to 1% for the different wedges. These uncertainties correspond to 1 SD of 5

measurements performed with the same diode on different days. For the EDP-10 diodes and

6 MV X-ray beams the dependence on field size of CF

wedge

is of the same order as the

uncertainty in the factor itself. Therefore, CF

wedge

is considered independent of field size

and CF

wedge

determined for a field size of 10x10 cm

is used.

In Table 5.2, the correction factors for the EDP-10, EDP-30, P30, QED and Isorad-p diodes

are shown for the different wedges. For the EDP-30, P30 and Isorad-p diodes, the mean of 5

measurements for three different diodes of the same type is given. In the case of the QED

diode, the correction factor is determined once for one diode.

Wedge angle (º)

wedge

EDP-10

1.009

1.013

1.018

1.035

wedge

EDP-30

1.002

1.004

0.998

1.009

wedge

P30

0.994

0.998

1.041

wedge

QED

1.005

0.999

1.012

1.015

wedge

Isorad-p

0.993

0.989

0.978

1.010

Table 5.2 CF

wedge

for different wedges for a 10 x 10 cm

field and for 6 MV X-rays (EDP-

10) or 18 MV x-rays (EDP-30, P30, QED and Isorad-p).

5.1.2.4 SSD CORRECTION FACTOR (CF

SSD

)

When the SSD is changed, the dose per pulse and the electronic contamination change.

First, the sensitivity of the diodes depends on dose per pulse. Secondly, if the build -up cap

of the diode is not thick enough an overestimation of dose at short SSD can be due to

electronic contamination that would be “seen” by the diode but not by the ionisation

chamber placed at the depth of dose maximum. Therefore, a SSD correction factor different

from 1 is expected.

The correction factor for SSD is defined as:

cm)

100

SSD

10x10

(

SSD)

10x10

(

cm)

100

SSD

10x10

(

SSD)

10x10

(

max

diodes

SSD

(10)

The diode is taped on the surface of a Plastic Water

phantom. The field size is fixed to 10

x 10 cm

at the isocentre. The reading of the diode measurement at different SSD normalised

to the reading of the diode measurement 100 cm SSD is compared to the same ratio

measured with an ionisation chamber placed inside a Plastic Water

phantom at the depth

of dose maximum.

In Figure 5.7

SSD

for different types of diodes and for an 18 MV X-ray beam is shown.

Figure 5.7 CF

SSD

for EDP-30, P30, QED and Isorad-p diodes for an 18 MV X-ray beam.

The mean and standard deviation for three diodes of each type and three

measurements are shown.

5.1.2.5 ANGLE CORRECTION FACTOR (CF

angle

)

To measure the directional dependence of the diodes they where placed with the

measuring point (considered to be at the b asis of the diode) at the isocentre on the surface

of the Plastic Water

phantom. The variation of the diode response with gantry angle for

a 10 x 10 cm

field was measured with the diode long axis perpendicular (axial symmetry)

and parallel (tilt symmetry) to the gantry rotation (Figure 5.8). The results of the

measurements are shown in Figures 5.9 and 5.10.

Figure 5.8 Design of a P30 diode. The plane containing x is the plane of gantry rotatio n

for tilt symmetry. The plane containing y is the plane of gantry rotation for

axial symmetry.

0.940

0.960

0.980

1.000

1.020

1.040

1.060

100

110

120

130

SSD (cm)

SSD

EDP30

P30

QED

Isorad-p

Figure 5.9 Axial and tilt symmetry for EDP-10 diodes, for 6 MV X-rays.

0.980

1.000

1.020

1.040

-80

-60

-40

-20

angle (°)

axial angle

EDP10

0.990

1.010

1.030

1.050

1.070

-80

-60

-40

-20

angle (°)

tilt angle

EDP10

Figure 5.10 Axial and tilt symmetry for EDP-30, P30, QED and Isorad-p diodes for 18 MV

X-rays.

0.980

1.000

1.020

1.040

1.060

1.080

-80

-60

-40

-20

angle (°)

axial angle

EDP30

P30

QED

Isorad-p

0.980

1.000

1.020

1.040

1.060

-80

-60

-40

-20

angle (°)

tilt angle

EDP30

P30

QED

Isorad-p

5.1.2.6 TEMPERATURE CORRECTION FACTOR (CF

temperature

)

To study the influence of temperature on the diode signal a water phantom equipped with a

thermostat is used. The diodes are taped on a thin slab of Perspex, which is in contact with

the water. The temperature, measured with a digital thermistor provided with an immersion

probe, is slowly increased from 22.5 ºC to 32 ºC. The sensitivity of the diodes is determined

at different temperatures and expressed relative to that at 22.5 ºC. Each temperature is

maintained approximately 20 minutes in order to reach full thermal equilibrium between the

surface of the phantom and the diodes. The procedure is performed twice, when the diodes

are received and after some time of use (1kGy of accumulated dose).

The temperature correction factor is defined as:

(

)

(

SVWT

temperatur

−

(11)

if the temperature at which the diodes have been calibrated (T

cal

) is 22.5°C.

SVWT is the sensitivity variation with temperature.

If T

cal

differs from 22.5ºC, the temperature correction factor is defined as:

(

)

( )

(

)

(

)

(

SVWT

cal

temperatur

−

(12)

For 10 EDP-10 diodes the variation of sensitivity (in percentage) per ºC varies from 0.26 to

0.34. Table 5.3 shows the sensitivity variation per ºC for different EDP-30 diodes, new and

after some time of clinical use. In Figure 5.11 the variation of sensitivity of EDP-30, P30,

QED and Isorad-p diodes is shown.

Diode number

new variation(%/ºC)

post-irradiation variation (%/ºC)

0. 290

0. 275

0. 293

0. 291

0. 320

0. 340

0. 216

0. 274

0. 281

0. 273

Table 5.3 Variation of sensitivity (SVWT) in % per ºC for EDP-30 diodes, when the diodes

are new and after some time of use.

Figure 5.11 Variation of sensitivity with temperature of (a) three EDP-30 and three P30

diodes, and (b) three QED and three isora d-p diodes. 21.5ºC was chosen as

normalisation temperature.

0.99

1.01

1.02

1.03

1.04

1.05

temperature t (ºC)

signal at tºC/signal at 21.5ºC

EDP30
P30

temperature t (ºC)

QED
isorad-p

5.1.2.7 INFLUENCE OF THE DOSE RATE ON THE DIODE’S SENSITIVITY

A Clinac 1800 (Varian) accelerator changes the dose rate by varying the number of pulses

per unit of time and not the dose per pulse. Therefo re, to test the intrinisic influence of

dose per pulse on the diode sensitivity, the following experiment was designed. In order to

exclude electron contamination, the diodes are inserted in a Plastic Water

phantom at 10

cm depth. Their flat surface is facing the beam. In this way, the measurements are not

affected by differences in build -up caps. The source-surface distance is then varied from 80

cm (0.56 mGy/pulse) to 130 cm (0.23 mGy/pulse). The field size is chosen constant and in

such a way that the phantom is completely irradiated at any distance. The test has been

performed for EDP-30, P30, QED and Isorad-p diodes. Results are shown in Figure 5.12.

In this experiment, no wedges are used to reduce the dose rate as by doing so, not only

dose rate would be modified but also the energy spectrum. The results of such an

experiment would mix up dose rate dependence with energy dependence.

Figure 5.12 Influence of dose rate on diode signal. Relative sensitivity is defined as the

ratio of the response of the diode at any dose per pulse to the response at the

dose per pulse corresponding to a SSD of 100 cm.

0.980

0.990

1.000

1.010

1.020

0.200

0.250

0.300

0.350

0.400

0.450

0.500

0.550

dose rate (mGy/pulse)

relative diode signal

EDP30

P30

QED

Isorad-p

5.1.2.8 SENSITIVITY VARIATION WITH ACCUMULATED DOSE (SVWAD)

The loss of sensitivity with accumulated dose has been studied for EDP-30, P30, QED and

Isorad-p diodes. Readings for 177 MU have been obtained after irradiations of 1500 MU (17

Gy) at 240 MU/min (dose rate used in clinical practice) in an 18 MV X-ray beam. A

monitoring ionisation chamber is used to avoid accelerator fluctuations influencing the

results of the study.

Figure 5.13 shows the loss of sensitivity with accumulated dose for the different diodes.

For

the EDP diodes the SVWAD can vary substantially. Much smaller values have been

reported [Meijer 2001], which are more in agreement with the specifications in Table 1.1.

Figure 5.13 Sensitivity loss with accumulated dose for EDP-30, P30, QED and Isorad-p

diodes

-8%

-7%

-6%

-5%

-4%

-3%

-2%

-1%

100

150

200

accumulated dose (Gy)

loss of sensitivity

EDP30

P30

QED

Isorad-p

Table 5.4 Summary of entrance correction factors for EDP-30, P30, QED and Isorad-p

diodes, for 18 MV X-rays. For EDP-30, P30 and Isorad-p diodes, the mean of

three diodes of each type and three determinations for each diode is given. For

QED diodes the mean of one measurement and three diodes is given, with the

exception of CF

wedge

and CF

angle.,

for which the results of one diode and one

measurement are given.

SSD = 100 cm

Open field

Field size (cm

)

5 x 5

10 x 10

15 x 15

20 x 20

30 x 30

EDP-30

0.978

1.000

0.993

0.984

0.968

P30

0.970

1.000

1.016

1.026

1.034

QED

0.986

1.000

1.006

1.008

1.011

Isorad-p

0.968

1.000

1.018

1.031

1.044

tray

SSD = 100 cm

Field size (cm

)

5 x 5

10 x 10

15 x 15

20 x 20

30 x 30

EDP-30

1.000

1.002

0.999

0.993

0.983

P30

1.001

1.000

0.998

0.996

QED

1.003

1.002

0.998

0.993

Isorad-p

0.997

0.999

0.998

0.997

SSD

Field size

= 10 x 10 cm²

Open field

SSD (cm)

100

110

120

EDP-30

0.973

0.992

1.000

1.010

1.021

P30

0.965

0.985

1.000

1.014

1.031

QED

0.964

0.987

1.000

1.012

1.025

Isorad-p

0.969

0.992

1.000

1.019

1.035

angle

(Axial symmetry)

SSD = 100 cm

Field size

= 10 x 10 cm

Open field

Angle

0º

10º

30º

45º

60º

EDP-30

1.000

1.002

1.022

1.042

1.050

P30

1.000

1.002

1.022

1.044

1.058

QED

1.000

1.002

1.020

1.033

Isorad-p

1.000

wedge

Wedge angle

15º

EDP-30

1.002

P30

0.994

QED

1.005

Isorad-p

0.993

SSD =100cm

Field size

= 10 x 10 cm

30º

45º

60º

1.004

0.998

1.009

0.998

1.041

0.999

1.012

1.015

0.989

0.978

1.010

SVWT

EDP-30

0.23-0.30

%/ºC

P30

0.15-0.21

%/ºC

QED

0.29-0.30

%/°C

Isorad-p

0.19-0.25

%/ºC

SVWAD

EDP-30

3.4%/100Gy

P30

0.2%/100Gy

QED

0.8%/100Gy

Isorad-p

0.3%/100Gy

Table 5.5 Summary of entrance correction factors for EDP-10 diodes for 6 MV X-rays. The

mean of ten diodes is given.

SSD = 100 cm

Open field

Field size (cm

)

5 x 5

10 x 10

15 x 15

20 x 20

30 x 30

40 x 40

EDP-10

1.000

0.996

0.992

0.987

0.985

tray

SSD = 100 cm

Field size (cm

)

5 x 5

10 x 10

15 x 15

20 x 20

30 x 30

EDP-10

1.002

1.000

0.996

0.992

0.987

SSD

Field size 10 x 10

Open field

SSD (cm)

100

EDP-10

0.982

0.991

1.000

angle

(Axial symmetry)

SSD = 100 cm

Field size 10 x 10

Angle

0º

10º

30º

EDP-10

1.000

1.002

1.008

5.2 PERFORMANCE OF SOME COMMERCIAL DIODES IN HIGH ENERGY

PHOTON BEAMS – THE LEUVEN EXPERIENCE

5.2.1 INTRODUCTION

For entrance dose measurements, in vivo diodes are covered with a build -up cap to enable

measurements at a certain depth. Firstly, in any case (unless for surface dose

measurements) some build -up is necessary in order to avoid the initial steep dose gradient.

Secondly, the build -up region is influenced by the variation in the amount of electron

contamination with different treatment geometries i.e. different source-surface-distance or

field size, or the presence of beam modifiers [Biggs 1979]. Depending on the cap thickness,

the diode will reflect this variation to a certain degree, in particular in high-energy photon

beams, where the influence of the contaminating electrons in the build -up region is larger,

mainly due to their increased range [Sjögren 1996]. Therefore, the build -up cap of the diode

should ideally have the same thickness as the build -up layer covering the ionization

chamber during the calibration, i.e. the build -up cap thickness should be equal to the depth

of dose maximum of the photon beam quality in use.

On the other hand, thicker build -up caps cause a larger perturbation of the treatment field

(which is however of only limited relevance if the diode is applied during only one or two

seesions). Commercially available diodes have different build -up cap thicknesses; some of

them are designed with a thinner cap to minimise field disturbance . With regard to the

diode correction factors to be used in non-reference conditions, this has important

consequences that may jeopardise the accuracy of in vivo dosimetry at high energies: i)

The magnitude of the correction factors will be larger. ii) In a clinical situation where it is

preferable to limit the number of factors, the factors have been established independently

and are used together for various combinations of beam setting. This may no longer be

straightforward if the degree of electron contamination, which will be depend on the

specific combination of accessories and beam set-up parameters, has a considerable

influence on the diode signal.

This problem is addressed in some recent studies performed in Leuven [Georg 1999],

Barcelona [Jornet 2000], Copenhagen [Lööf 2001], and Amsterdam [Meijer 2001]. Only

limited information is available describing the correction factor variation and/or the

achievable accuracy for in vivo dosimetry methods in the ‘high’ energy range (16 – 25

MV). Therefore, the first aim of these studies was to assess and analyse the variation of

diode correction factors for entrance dose measurements at higher photon beam energies.

For the investigations performed in Barcelona and Copenhagen commercially available n -

type as well as p-type diodes have been included. The dosimetric characteristics of n -type

diodes have been published in [Jornet 2000]. The results of the comparative study of EDP-

30, QED (p-type diodes) and P30 (n -type diode) in Barcelona are shown Section 5.1, and are

substantiated by the results obtained in Copenhagen [Lööf 2001] and Amsterda m [Meijer

2001].

Results of the study performed in Leuven ([Georg 1999]) will be presented in this

contribution. In addition to determining the correction factors for commercial diodes, the

total build-up thickness of the diodes is modified in order to min imise the correction factor

variation.

5.2.2 MATERIAL AND METHODS

5.2.2.1 MATERIAL

The Scanditronix p-type diodes that are recommended for in vivo dosimetry in high energy

photon beams have been investigated: two different EDP-20 diodes - “old” and “new” type

- and EDP-30 diodes. The specifications of these diodes are listed in Section 1.1.4, Table

1.3. The main difference between old and new EDP-20 diodes is the higher doping level of

the new type, which was introduced in order to improve the dose rate properties with

accumulated dose. The EDP-30 also has a high doping level

Furthermore the new EDP-20

type is covered with a thin (0.2 - 0.3 mm) plastic layer. The old type EDP-20 diode, which is

no longer produced, is included in this study since it is still in clinical use.

100

For all measurements diodes are connected to commercially available or home made

electrometers with a low input impedance [Rikner 1987]. Additional build -up caps have

been manufactured in the hospital’s mechanical work-shop. These build -up caps have

(nominally) thickness equivalent in attenuation to 10, 15, 20 and 30 mm water, and are made

from either copper, lead, or iron. The build -up caps can easily be added to and removed

from the diode without any damage. Whenever modified with an additional build -up cap,

the diode is described by adding as an index the total build -up thickness of the (modified)

diode, and the build -up cap material is indicated by its chemical symbol, e.g. measurements

made with an EDP-20/30(Cu) diode indicate that the EDP-20 diode has been modified by a

copper build -up cap equivalent in thickness to 10 mm water. For EDP-30, the thickness of

the commercial (Tantalum) build -up cap corresponds to approximately 14 mm water

equivalent material [Jornet 1996 and Section 5.1.1.4]. Although this value is apperently not

fixed, but depends on the head-scatter spectrum of the accelerator [Meijer 2001, Sjögren

1998], we have assumed it to be around 15 mm. Therefore, e.g. EDP-30/30(Cu) means that

an additional copper build -up cap equivalent in thickness to 15 mm water has been added.

Ionisation chamber measurements are performed using a cylindrical ionisation chamber (NE

2571, volume 0.6 cm

). All diode correction factors are determined in a large polystyrene

phantom consisting of plates of different thicknesses.

The diode characteristics are investigated in 18 MV, 23 MV and 25 MV photon beams,

provided by different linear accelerators: a Philips SL20 (18 MV, QI = 0.78), a GE Saturne

43/Series 800 (18 MV, QI = 0.77, 25 MV, QI = 0.786), two GE Saturne 42/Series 700 (18 MV,

QI = 0.77 and 0.78), and a GE Sat II+ (23 MV, QI = 0.79). The GE Sat 43/800 is equipped with

an integrated multileaf collimator wh ile all other machines have conventional collimators.

The main difference between the Philips and the GE accelerators is the treatment head

geometry, especially the collimator design and the flattening filter material and position.

The influence of a wedge filter on the different diode types has been investigated in the

beams of the Saturne 43 linac. The internal tungsten alloy wedge (nominal wedge angle

102

5.2.2.2 METHODS

The entrance dose is measured with the diodes using the methodology described in

Section 1.2 and Section 5.1. Field size correction factors are assumed to be independent of

SSD and SSD correction factors are assumed t o be independent of field size:

C(FS, SSD)

≈ C

(FS, 100 cm) • C

SSD

(10 x 10 cm

, SSD).

(13)

In order to verify equation (13) at higher photon energy, where electron contamination can

have a significant influence, several measurements are performed at 80 and 120 cm SSD for

large and small field sizes (5 x 5 cm

and 30 x 30 cm

), at 18 and 25 MV. If this assumption

would not be valid in the energy range under consideration, a two -dimensional correction

factor table should have to be determined, requiring tedious measurements.

Assuming the validity of equation (13), field size correction factors C

are measured for

square fields ranging from 5 x 5 cm

to the maximum field size of 40 x 40 cm

at reference SSD

(100 cm). Source-surface-distance correction factors C

SSD

are measured for SSDs varying

from 80 to 120 cm for the reference field size (10 x 10 cm

Wedge correction factors C

wedge

are measured at reference SSD for square fields of 5 cm, 10

cm and 20 cm field size. Since the wedge position is critical, measurements with wedge are

performed for collimator orientations of 90

and 270

, and C

wedge

is determined as the

average value. A tray correction factor C

tray

is determined repeating all C

and C

SSD

measurements for a block tray. Block correction factors C

block

are measured for a fixed

collimator opening of 20 x 20 cm

for blocks defining square fields of 5 cm, 8 cm, 10 cm, 14.1

cm, and 17.3 cm side length (at the isocenter plane). The block-to-isocentre distance varies

between 32 and 38 cm for different linacs.

The variation of correction factors for diodes from the same batch is estimated from C

and C

SSD

measurements for three EDP-30 and two new type EDP-20 diodes.

103

5.2.3 RESULTS

5.2.3.1 INDEPENDENCE OF FIELD SIZE AND SSD CORRECTION FACTORS

Equation (13) has been checked for old and new type EDP-20 and EDP-30 diodes in 18 MV

and 25 MV photon beams provided by GE linacs (SAT 42 and 43). The agreement between

the measured correction factor C(FS,SSD) and the correction factor calculated from C

and

SSD

is between 1 - 1.5 % for modified and unmodified old type and new type EDP-20

diodes. The larger deviations around 1.5 % are observed for a 5 x 5 cm

field size either at 80

or at 120 cm SSD. For the modified EDP-30/30(Cu) diodes deviations do not exceed 1.5 %,

while for unmodified EDP-30 diodes maximum deviations around 2 % are observed for a 5 x

5 cm

field at 80 cm SSD. Total build -up thickness larger than 30 mm (e.g. old type EDP-

20/40(Cu) or EDP-30/35(Cu)) did not improve these results.

Centre

Charleroi

Leuven

St. Jean

Unit

Ph SL 20 GE Sat 43 GE Sat 42 GE Sat 42

FS [cm]

QI=0.78 QI=0.77

QI=0.77

QI=0.78

5 / 6

1.003

1.006

1.013

1.000

0.994

0.989

0.990

0.981

0.986

0.981

0.979

0.970

0.987

0.973

0.972

0.963

0.986

0.962

0.960

0.955

Table 5.6 Field size correction factor variation for unmodified old type EDP-20 diodes at

18 MV photon beams measured in different Belgian radiotherapy centers.

104

5.2.3.2 FIELD SIZE CORRECTION FACTOR C

WITHOUT TRAY

Field size correction factors for unmodified old type EDP-20 diodes are shown in Table 5.6

as a function of square field size for 18 MV photon beams provided by different linacs. For

field sizes between 5 x 5 cm

and 30 x 30 cm

varies by only 1.7 % on the Philips SL 20

linac, and around 5 % on GE linacs.

Old type EDP-20 diodes without additional build -up show the largest field size correction

factor variation. In a field size range between 5 x 5 cm

and 40 x 40 cm

varies by about

4-6 % in 18 MV and 25 MV photon beams provided by the GE Sat 43 and 42 accelerators.

When adding build -up caps, the variation could be substantially reduced. For example, for

modified old type EDP-20/30(Cu), for the same field size range and energies, C

varies

only between 1 - 1.5 %.

New EDP-20 diodes without additional build -up show a much smaller C

variation than

the old ones: 2.2 and 2.5 %, respectively, in 18 MV and 25 MV beams provided by the Sat

43 linac. Only a slightly smaller variation of 2 % could be obtained at 18 MV for the

modified new type EDP-20/30(Cu), but the C

variation of 2.5 % at 25 MV could not be

improved when modifying the build -up thickness for this type of diode. For both types of

EDP-20 diodes the C

variation increased when a total water equivalent build -up

thickness of more than 30 mm is used.

EDP-30 diodes without additional build -up show a larger variation of C

than new type

EDP-20 diodes. When increasing the FS from 5 x 5 to 40 x 40 cm

the difference between

maximum and minimum C

values reaches 4 % and 4.4 % at 18 MV from the Sat 42 and 43,

but is only 2.4 % at 25 MV. For the EDP-30/30(Cu) the C

variation decreases to 0.5 % for

the 18 MV beam of the Sat 42 accelerator and to less than 1.5 % for the 18 MV beam of the

Sat 43 accelerator, but it increases slightly at 25 MV. For the EDP-30/35(Cu) and EDP-

30/45(Cu) the C

variation increased as compared to the EDP-30/30, this increase is more

pronounced at 25 MV. Figure 5.14 shows the variation of field size correction factors as a

function of field size, dependent on build -up thickness for the different diodes in a 18 MV

photon beam provided by the GE Sat 43 accelerator.

105

Figure 5.14 Variation of the field size correction factor C

for different types of diodes

with and without additional build-up in 18 MV photon beams provided by a

GE Sat 43 linac: (a) old type EDP-20, (b) new type EDP-20, (c) EDP-30. All

results are obtained with copper build-up caps. The error bars indicate the

accuracy of

± 0.5 % in the determination of diode correction factors.

0.95

0.96

0.97

0.98

0.99

1.00

1.01

1.02

NEW EDP-20

NEW EDP-20/30

NEW EDP-20/40

(b)

0.95

0.96

0.97

0.98

0.99

1.00

1.01

1.02

OLD EDP-20

OLD EDP-20/30

OLD EDP-20/40

(a)

106

5.2.3.3 SSD CORRECTION FACTOR WITHOUT TRAY

Table 5.7 shows the variation of C

SSD

for different linacs fo r unmodified old type EDP-20

diodes. For the same quality index the difference in SSD correction factors (at a given SSD)

reaches 1 % at maximum at small SSDs. When decreasing the SSD from 100 to 70 cm, C

SSD

decreases by about 7 % for 18 MV photon beams with a QI = 0.78, and around 10 % for a

QI = 0.77.

In 18 MV and 25 MV photon energies provided by the GE Sat 43 and/or Sat 42 linacs, SSD

correction factors without tray are almost independent of build -up cap thickness for all

types of diodes. If the total build-up cap thickness does not exceed 30 mm, C

SSD

for the

same diode with and without additional build -up varies by less than 1 % at a given SSD.

Maximum deviations of more than 1.5 % between C

SSD

of unmodified and modified diodes

(at a given SSD) are observed only if the total build -up thickness reaches 40 or 45mm.

Centre

Charleroi

Leuven

St. Jean

Unit

Ph SL 20 GE Sat 43 GE Sat 42 GE Sat 42

SSD [cm] QI=0.78

QI=0.77 QI=0.77

QI=0.78

100

1.000

0.981

0.975

0.976

0.980

0.960

0.940

0.941

0.958

0.95

0.96

0.97

0.98

0.99

1.00

1.01

square field size (cm)

EDP-30
EDP-30/25
EDP-30/30
EDP-30/35
EDP-30/40

107

0.930

0.898

0.908

0.923

Table 5.7 Variation of SSD correction factor for unmodified old type

EDP-20 diodes at 18

MV photon beams measured in different Belgian radiotherapy centers.

Figure 5.15 Variation of the source-surface correction factor C

SSD

with additional

build-up for new and old type EDP-20 diodes in 18 MV photon beams

provided by a GE Sat 43 linac. All results are obtained w ith copper build-

up caps.

The new type of EDP-20 diodes shows a much smaller variation of C

SSD

with SSD than the

old type. For new type EDP-20 diodes C

SSD

increases by about 4-5 % at 18 and 25 MV from

GE linacs when increasing the SSD from 80 cm to 120 cm. This variation exceeds 10% for

the old type EDP-20 for almost all build-up cap combinations. Figure 5.15 shows the C

SSD

variation as a function of source-surface distance and build -up for new and old type EDP-

0.95

0.96

0.97

0.98

0.99

1.00

1.01

1.02

1.03

1.04

1.05

100

110

120

130

SSD (cm)

SSD

OLD EDP-20

OLD EDP-20/30

OLD EDP-20/40

NEW EDP-20

NEW EDP-20/30

NEW EDP-20/40

108

20 diodes in 18 MV photon beams provided by a GE Sat 43 linac. The error bars indicate the

accuracy of

0.5 % in the determination of diode correction factors. All results of modified

diodes displayed in Figure 5.15 are obtained with copper build -up caps.

For EDP-30 diodes with a total build -up less than or equal to 30 mm, C

SSD

increases by

about 3.5 - 4.5 % at 18 and 25 MV when increasing the SSD from 80 cm to 120 cm. In 25 MV

photon beams this variation is only 2% for the modified EDP-30/45(Cu) diode.

5.2.3.4 INFLUENCE OF BEAM MODIFIERS: TRAY AND BLOCK CORRECTION

FACTOR C

AND C

Without additional build -up C

values for a 40 x 40 cm

field with and without tray show

variations up to 2% for EDP-20 and 3% for EPD-30 diodes. This difference decreases to

about 1 - 1.5 % for a 30 x 30 cm

field and to less than 1 % for fields up to 20 x 20 cm

For unmodified diodes the SSD correction factor with and without tray varies by about 1.5

–2 % at a SSD of 80 cm. This variation is almost independent of energy and t ype of diode.

For SSDs between 90 and 120 cm the influence of the tray on C

SSD

is less than 1%.

The influence of the tray at small source-surface distances and for large field sizes can be

decreased by adding build -up. For modified EDP-20/30(Cu) and EDP-30/30(Cu) diodes

correction factors with and without tray, field size correction factors C

as well as source-

surface correction factors C

SSD

, differ by less than 1%.

Table 5.8 shows the variation of block correction factors as a function of blocked field size

for different types of diodes and linacs. The collimator setting is kept constant at 20 x 20

for all blocked fields. The build -up material is indicated by its chemical symbol.

For unmodified old type EDP-20 diodes block correction factors at 18 MV provided by the

Philips SL 20 linac do not differ by more than 1 % from unity, even when reducing the field

size to 5 x 5 cm

for a fixed collimator setting of 20 x 20 cm

. For the GE Sat 42 linac, C

reaches 1.032 at 18 MV for t he smallest field size. If sufficient build -up material is provided,

109

the influence of blocks can be significantly reduced. For the modified EDP-20/30 at 18 MV

from the GE Sat 42 C

could be reduced to 1.4% for the blocked 5 x 5 cm

5.2.3.5 WEDGE CORRECTION FACTORS

Wedge correction factors at 18 and 25 MV are of the order of 1.01 - 1.02 for new type EDP-

20/30(Cu) and EDP-30/30(Cu) diodes, and reach 1.06 - 1.07 for the old type EDP-20/30(Cu)

diodes.

The variation of the wedge correction factor C

with field size is less than 1 % for modified

and unmodified new type EDP-20 and EDP-30 diodes at 18 and 25MV. Only for unmodified

old type EDP-20 diodes a small field size dependence could be observed: the C

difference

between a 5 x 5 cm

and a 20 x 20 cm

field reaches 2 % at 25 MV and is slightly higher than

1 % at 18 MV. With additional (copper) build -up (old type EDP-20/30 or EDP-20/40) this

variation is reduced to less than 1 %.

Centre

Charleroi

Leuven

St. Jean

Middelheim

Unit

Energy

Ph SL 20

18MV

0.78

Sat 42

18MV

0.77

Sat 42

18MV

0.77

Sat II+

23MV

0.79

FS [cm]

EDP-20

EDP-20/30 (Fe)

EDP-20

5.0

1.009

1.032

0.986

1.032

8.0

1.007

1.032

0.993

1.028

10.0

1.007

1.026

0.997

1.025

14.1

1.000

1.012

0.999

1.017

17.3

1.003

1.007

1.012

20.0

1.000

110

Table 5.8 Variation of block correction factor for several unmodified and modified

diodes for higher energy photon beams measured in different Belgian

radiotherapy centres. The results of EDP-20 diodes refer to the old type.

5.2.3.6 CORRECTION FACTOR VA RIATION WITHIN THE SAME BATCH

The correction factor variation for diodes from the same batch is estimated from

measurements using unmodified diodes, which show the largest C

variations. The C

difference for a given field size is less than 1 % for two EDP-20 diodes and reaches 1.3 % as

a maximum for the three EDP-30 diodes. The difference of SSD correction factors for a fixed

SSD is less than 1 %, for both types of diodes. The small deviation of correction fac tors for

diodes from the same batch is in agreement with observations from other authors [Fontenla

1996b]. It can therefore be concluded that it is sufficient to determine correction factors for

one diode of the same batch only.

5.2.3.7 PERTURBATION EFFECTS

Perturbation effects are determined for old and new type EDP-20 and EPD-30 diodes with

and without additional copper build -up caps in a 10 x 10 cm

field at 18 and 25 MV. The

relative dose reduction (dose reduction in % with respect to the flat part of the profile at a

specific depth) for unmodified old and new type EDP-20 diodes is about 7-8 % at 18 MV,

and 6-6.5 % at 25 MV, respectively. The corresponding values for the unmodified EDP -30

diode are between 3 % and 3.5 % at 18 and 25 MV, respectively. Depending on energy and

depth, these values are increased by 2-3 % when adding copper build -up caps.

The additional build -up caps have thickness typically around one mm and increase

therefore the area where perturbation effects are present.

111

5.2.4 DISCUSSION

5.2.4.1 INDEPENDENCE OF FIELD SIZE AND SOURCE-TO-SURFACE DISTANCE

CORRECTION FACTORS

For SSDs larger than or equal to 80 cm equation (13) is valid within an uncertainty of 1.5 %

at 18 - 25 MV if the total build -up cap thickness does not exceed 30 mm. It should be noted

that the accuracy in the determination of diode correction factors is better than ± 0.5 %.

The validity of equation (13) within a certain limit has to be considered when defining

tolerance and action levels for entrance dose measurements in higher energy photon

beams, especially at small or large SSDs.

The head-scatter variation of a therapy unit depends strongly on treatment head geometry

([Dutreix 1997], [Nilsson 1998], [Sixel 1996], [Van Gasteren 1991]), and GE linacs are known

to have a pronounced head-scatter variation with field size. Therefore, the independence of

and C

SSD

was checked on two GE linacs. Since field size correction factors vary much

less for other linac types (see Table 5.7) the findings for the GE linacs are considered to be

valid for other linacs, too.

5.2.4.2 TOTAL BUILD-UP THICKNESS OF THE DIODE

For EDP-30 diodes the smallest variation of the field size correction factor at 18 MV is

obtained with an additional water equivalent (copper) build -up thickness of 15 mm. The

results for both types of EDP-20 diodes are in good agreement with this. At 18 MV, as well

as at 25 MV, for old and new EDP-20 diodes the smallest FS correction factor (C

)

variation is observed when adding 10 mm water equivalent build -up material, which also

corresponds to about 30 mm depth. This optimal depth has been confirmed using different

build-up cap materials: copper, lead and iron. It should be noted that even with a modified

EDP-10/30(Cu) diode a 1.6 % C

variation is seen at 18MV for the Ph ilips SL 20 linac when

the field size varies between 6 x 6 and 35 x 35 cm

. For EDP-30 diodes in 25 MV photon

112

beams, the C

variation of unmodified diodes could not be decreased by adding build -up.

The different diode response at 25 MV for EDP-20 and EDP-30 diodes is most probably due

to the difference in build -up material of the unmodified diodes (stainless-steel versus

tantalum).

5.2.4.3 TREATMENT UNIT DEPENDENCE

For the same diode or diodes from the same batch, field size and source-surface distance

correction factors measured at 18 MV on the GE Sat 43, equipped with an MLC, and

measured on the GE Sat 42, having a conventional collimator (CC), agree mostly within 1%.

The main difference between the two linacs are the collimators, while the flattening filter

position and design are similar. Due to the different collimator design they show a

difference in head-scatter variation with field size (12 % for the GE Sat 43+MLC linac versus

16 % for the Sat 42+CC linac when increasing the field size from 4 x 4 to 40 x 40 cm

), but the

depth dose characteristics in the build -up region are almost identical. This finding is in

agreement with a study where two different collimators have been mounted subsequently

on the same accelerator, while all other parts of the linac rema ined the same [Georg 1997].

Here, it has been shown that the depth of maximum dose as well as the skin dose is almost

independent of collimator design while the head-scatter variation is not. Similar doses in

the build -up region but different head-scatter variation may be explained by contaminating

electrons emanating from the flattening filter, which is the main source of contaminating

electrons at higher energies ([Nilsson 1985], [Nilsson 1986], [Sjögren 1996]). The head -

scatter variation, on the other hand, has to be determined at depths large enough to avoid

the unpredictable influence of contaminating electrons in order to describe the field size

variation of the primary photon component [Dutreix 1997].

The same arguments can be used to explain the difference in diode correction factors for

the same type of diode for the same photon beam quality but provided by linacs from

different manufacturers.

113

5.2.4.4 BEAM MODIFIERS

Even for large field sizes at small SSDs the influence of the tray could be decreased to

negligible values when adding additional build -up caps. It should be noted that the

thickness of the tray holder is only 8 mm; thicker trays may have a larger influence.

Additional build -up material also reduces the influence of blocks to less than 1% for

clinical relevant applications. The influence of trays and blocks can be explained by

contaminating electrons. At higher energies the dose contribution of contaminating

electrons produced by trays and blocks is only about 5 – 10 % at depths around 3 - 4 cm,

but between 20 and 40 % at depths up to 2 cm ([Bjärngard 1995], [Sjögren 1996], [Zhu

1998]).

Block correction factors for modified diodes have to be taken into account only when the

field size is substantially reduced by blocks (see Table 5.8). Large field reduction by blocks

can reduce the area of flattening filter seen by the detector which might be accounted for

by a field size correction factor rather than by a block correction factor. However, such

block applications are clinic ally less relevant.

The different doping level for the new type EDP-20 diode was introduced in order to

improve the dose rate linearity of the diode after dose accumulation [Grusell 1993].

Therefore the new type EDP-20 diodes show a smaller variation of SSD correction factors

as well as a smaller wedge correction factor.

5.2.4.5 PERTURBATION EFFECTS

The relative dose reduction caused by the diode is almost independent of depth, which is

in agreement with investigations performed with diodes from another manufacturer [Alecu

1997]. The attenuation at higher photon energies induced by unmodified EDP -20 diodes is

somewhat larger than previously reported values for Scanditronix diodes ([Leunens 1990b],

[Nilsson 1988], [Rikner 1987]), while it is smaller for the EDP-30. When adding build -up

caps these perturbation effects are increased for all diodes, but they are still smaller than

114

the 13 % attenuation in a 15 MV photon beam reported for a diode with a cylindrical build -

up cap [Alecu 1997] and Section 5.1.1.

When performing in vivo dosimetry on a weekly basis and assuming a perturbation effect

of 10 %, the dose reduction due to the presence of the diode will be 2 % for treatments with

5 sessions per week. This value can even be reduced by care fully varying the diode

position for each in vivo measurement or by restricting the in vivo procedure to the

beginning of a treatment.

5.2.5 CONCLUSION

When performing entrance dose measurements at the depth of maximum dose, the diode

correction factors accounting for non-reference conditions strongly depend on

contaminating electrons, whose dose contributions at shallow depths vary with treatment

geometry and treatment unit. The build -up thickness of commercial diodes is not sufficient

to exclude this influence at higher photon energies. In the energy range between 18 - 25

MV a total diode build -up thickness of 30 mm is found to be the ‘best compromise’ for all

types of diodes and treatment units considered. An additional advantage of a build -up cap

modification when using commercial diodes at higher energies, is the reduction of the

influence of beam modifiers. The additional perturbations caused by the increased build -up

thickness do not have an influence on the practical aspects of in vivo dosimetry (for

instance the frequency of measurements) compared to unmodified diodes.

115

5.3 PRACTICAL IMPLEMENTATION OF COST-EFFECTIVE APPROACHES TO

IN VIVO DOSIMETRY - THE EDINBURGH EXPERIENCE

5.3.1 INTRODUCTION

A systematic programme of in vivo dosimetry using diodes to verify radiotherapy delivered

doses was begun in Edinburgh in 1992. Prior to that, TLD -based in vivo dosimetry had

been used for some considerable time, but only for critical organ dosimetry and for

verification of complex treatments, such as TBI and TSEI. [Thwaites 1990] This has

continued. In 1991, a Scanditronix DPD-6 electrometer and a set of p-type silicon diodes

(EDE, EDP-10, EDP-20) was purchased with the following initial aims:

•

to investigate the feasibility of routine systematic use of diodes as part of a

comp rehensive QA programme

•

to carry out clinical pilot studies to test each machine, treatment site and treatment

technique to assess the accuracy of the overall radiotherapy process at the point of

dose delivery

•

from this, to identify and rectify any systematic errors, and also

•

to measure global (as well as specific treatment machine and technique) dosimetric

precision to compare to clinical requirements

As the programme progressed, further electrometers (DPD3 and DPD510) and diodes

(EDD2, EDD5, newer-style EDP-10 and EDP-20, all Scanditronix) were acquired and

additional aims were developed linked to routine implementation. From the results of the

clinical pilot studies, a cost-benefit evaluation was carried out to consider how best to

utilise the diodes in ro utine practice, including consideration of:

•

when to use and on which patients

•

what should be measured in a routine programme

•

who should do what

•

what tolerance and action levels to adopt

•

what action should be taken

116

At the same time, further work was initiated to attempt to simplify the routine

implementation to make it as cost-effective as possible. This has included:

•

comparison of the use of correction factors versus no correction factors

•

the use of additional build -up caps on the standard diodes to minimi se the ranges of

correction factors required

•

the use of combined mid -range ‘generic’ correction factors for a specific modality,

treatment site/field and diode position

•

how data is communicated and recorded, to and from the treatment unit

•

how the diodes are mounted and handled in the treatment room

•

the quality control required for the diodes themselves

The department was fortunate in having a steady supply of multi-disciplinary Master’s

level students interested in the work, who carried out dissertation projects at the testing

and development stages under the supervision of the physicist in charge of the diode

project. Trainee physicists have investigated the initial physics testing and workup of the

systems, including calibration and corrections of the dio des ([Kidane 1992], [Brown 1996])

and phantom and clinical pilot studies [Brown 1996]. A part -time (one day per week)

research radiographer, working within the physics department has been involved with the

project since soon after it started and has carrie d out work on initial clinical implementation,

clinical pilot studies, routine implementation and methodology ([Blyth 1997], [Blyth 2001]).

Trainee radiation oncologists have carried out clinical pilot studies ([Millwater 1993],

[Millwater 1998], [Elliott 1999]). At the time of writing, a number of papers are in

preparation, or submitted, covering various aspects of this programme ([Blyth 2001a],

[Blyth 2001b], [Elliott 2001]).

Implementation on the linacs in the department has been gradual. Initial detail ed studies

were carried out on just one linac, then this has been extended one-by-one to all linacs in

the department. Routine implementation has also been rolled out gradually. Thus different

117

stages of the programme have been in operation on different lin acs at the same time. This

process is not yet complete. At the time of writing, the department is undergoing a major

re-equipping. Allied to this, new diode mounts are being installed in each room, as new

linacs are installed, to enable routine use on all treatment machines.

The following sections briefly discuss the approaches which have been taken in

development and those adopted for routine usage in the department, with an emphasis on

cost-effective, reliable, readily -usable methods. It should be noted that the solutions

chosen are specific to the department. The optimum solutions are not necessarily the same

for every department (nor even for every treatment type within a given department), as

they are influenced by the resources available, the level of implementation, the accuracy

deemed necessary, the other aspects of the department’s quality system and QA

programme in operation, etc. However it is observed that many of the solutions have been

arrived at independently by other departments who have gone through a similar process,

including others contributing to this booklet. Where details are similar and are discussed in

other chapters or sections, they will not in general be repeated here.

5.3.2 INITIAL PHYSICS TESTING AND WORKUP

A conventional approach wa s taken to the initial physics testing and workup. Whilst this

is time-consuming, it is necessary to carry this out in full detail at the outset, whatever the

level of routine implementation is to be. Confidence in the diode programme results, in the

evaluation of clinical pilot studies and in the tolerance levels applied can only be based on

a comprehensive commissioning and evaluation of the diode systems. Thus initial testing

included (see Sections 1.2.1, 5.1): stability of diode signal (leakage), reproducibility of

system response to repeated irradiations, measurement linearity, check on water-equivalent

depth of the diodes and measurement of the related perturbation of the radiation field

beyond the diodes.

118

Following this, entrance and exit dose calibrations were carried out (see Section 1.2.2 and

[Millwater 1997]), comparing diodes to a calibrated ion chamber irradiated simultaneously.

The calibration phantom used wa s of 30 x 30 cm

epoxy -resin water-equivalent plastic and a

standard thickness of 15 cm was chosen. The phantom was set up over the thin meshed

area of the treatment couch. For entrance dose calibration, the ion chamber was positioned

at the depth of dose maximum and an SSD of 100 cm was set. To minimise subsequent

routine calibration times, the methodology tested and implemented was to position a

number of diodes spaced around the central axis for calibration, sufficiently far out so as

not to perturb the ion chamber reading. A field size of 15 x 15 cm

was selected, to ensure a

sufficient field margin around this ring of diodes. The dose calculated from the calibrated

ion chamber was corrected for the displacement factor, whilst the diode readings were

corrected for the small measured beam non-flatness at their distance out from the central

axis. For exit dose calibrations, the ion chamber and diodes were left in the same positions,

the gantry rotated to 180o, the SSD was reset to 100 cm and simultaneous irradiations were

again carried out. As the exit diode is to be used to compare measured doses against those

calculated using local planning data, the exit calibration factors were corrected for the

measured lack of scatter to the ion chamber at its exit calibration depth (build -down effect).

Entrance and exit correction factors were determined for each individual diode. Standard

methods were used for this (e.g. Sections 1.2.3 and 5.1; [Van Dam 1994], [Mayles 2000]) and

similar magnitude corrections were obtained to those reported in the literature. Correction

factors were determined for each beam for field size, SSD, tray, wedge, blocks, incident

angle and temperature. All were investigated for both entrance and exit measurements. In

addition for exit measurements, the correction factors for phantom thickness were

determined.

Following all of this initial work, irradiations were carried out on phantoms, comparing

diode-measured doses to expected doses in a variety of situations to test the applicability

119

of the methodology and the validity of the calibration and correction factors. In addition

detailed phantom studies have been carried out to aid in relating entrance and exit dose

measurements to isocentre dose estimation, in order to compare measured doses to

prescribed target volume doses.

At this stage, a quality control programme was implemented for the diodes. Initially this

included checks on calibrations approximately monthly and on correction fac tors

approximately yearly [Mayles 2000]. In addition records were begun of the approximate

cumulative doses that each diode had received, as an indication of whether checks should

be more frequent. It was recognised with experience that calibration factors can be

routinely and frequently checked in practice by using the diodes in the linac daily check

procedures, building in a quick consistency check against an ion chamber on at least a

weekly basis. A quick check on correction factor validity can be carrie d out using a second

SSD measurement (see Section 1.2.4). If any of these show significant changes then this

indicates the need for more detailed checks.

5.3.3 PILOT CLINICAL STUDIES

Pilot clinical studies were carried out for diffe rent treatment machines, beams, treatment

sites and treatment techniques. Simpler situations were investigated first, e.g. flat surfaces,

shells, perpendicular incidence, etc., where diode positioning problems were expected to be

less, whilst the underlyin g methodology was tested in the clinical setting. Entrance and exit

doses were measured once per week throughout treatment. All relevant correction factors

were applied to the measurements. The only exception to this was the temperature

correction, as the diode was generally positioned after field setup and just before

irradiation, therefore it was felt in most circumstances that the correction required due to

the temperature change was minimal. Temperature corrections were applied if the diode was

in posit ion for a significant

time. Diodes were positioned wherever possible on the beam

central axis, but account was taken of block presence, asymmetric field position etc. where

appropriate. The doses were compared to expected entrance and exit doses calculated from

120

the planning data, and the deviations were quantified. These were used to estimate target

volume prescription point dose deviations from the individual fields, aided by the phantom

studies. The values for all fields were combined in proportion to the weight of the field to

provide an estimate of the overall deviation of delivered dose from prescribed dose for the

whole treatment. For tangential field breast treatments, the diodes were placed at a point

midway between the field centre and the medial border of the medial field and used there to

measure entrance dose from the medial field and exit dose from the lateral field. The

combined corrected dose was compared to the dose value from the plan at the appropriate

depth below this measurement point.

In all the clinical pilot studies measurements were repeated on a weekly basis throughout

treatment. The observed deviations from all measurements on the same patient were

averaged to provide the best estimate of the overall treatment course deviation betwee n

delivered dose and prescribed dose to the target volume prescription point.

Detailed results are discussed elsewhere, but in summary:

•

Typical distributions of individual entrance dose results for various clinical pilot studies

showed mean differences, measured to expected, close to zero. Standard deviations lay

within the range 1.2 % to 4.1 % (typically 1.5 – 3 %), depending on site and linac.

•

Exit doses showed mean (systematic) differences varying with site, field and method of

treatment planning. Typic ally mean measured doses were observed to be lower than

expected by 1 – 4 %. Standard deviations were within the range of 1.9 – 5 % (typically 3 -

4.5 %)

•

For total treatment course dose delivery to the target volume (prescription point), some

examples of mean deviations (followed by SD), from the combination of all fields and

using all information from repeated measurements are:

head and neck:

-0.2 to +1.0 % (1.5 – 3 %)

121

breast:

-4 %

(2.5 %)

(old technique, isocentre on surface, half beam in air [Redpath 1992])

breast:

-2 %

(2.7 %)

(new technique, isocentre at depth, less beam in air [Carruthers 1999])

pelvic:

-0.4 %

(2.7 %)

conformally blocked prostate and bladder(initial):

+1.5 %

(2.6 %)

conformally blocked prostate and bladder(corrected):

+0.1 %

(2.6

The SD for the distribution of observed differences of estimated prescription point doses

(from combined field measurements) from the expected values was normally lower than the

SD for individual field data. Similarly the SD for measurements which were repeated on the

same patient and averaged over the treatment course was lower than the s.d for individual

measurement data. Some of the reasons for observed discrepancies between measured and

expected doses, on investigation, were found to be due to the in vivo methods. These

include diode positioning problems such as contact, cable pulling, etc.; positioning

difficulties such as entrance or exit through couch, etc.; wrong correction factor use;

measurements through immobilisation devices, etc.; the limiting resolution of the diode

electrometer for small dose wedged components of fields; diode positioning uncertainties

under large wedges or on steeply angled surfaces. On the other hand some causes were

identified as real differences in delivered dose, due to treatment machine performance, to

patient data acquisition, to dose calculation errors (e.g. for tangential field breast

irradiation), to the use of non-CT planning in some situations, to patient set-up variations,

and to incorrect treatment parameters. Some causes were due to changes in the patient at

the time of treatment as compared to the plan, for instance systematic patient size and

shape changes, or random changes such as bowel gas in line with the diode, etc. Some

causes were a combination of factors. The conformal blocked treatments illustrate one such

case. The initial mean deviation was observed to be + 1.5 % (apparent measured dose

greater than expected). On investigation approximately half of this difference was due to an

122

incorrect correction to the diode reading to account for the presence of the conformal

blocks (i.e. diode use and methodology) and approximately half was due to the MU

calculation in these situations (i.e. real change to delivered dose to the patient). On

correcting both of these errors the mean deviation subsequently measured was 0.1 %. This

example nicely illustrates that all discrepancies should be investigated, that diode

dosimetry is precise enough to identify problems at the 1 %, or sub-1 %, level if the system

is implemented carefully and that the diode methodology and use can itself introduce

errors, which should also be suspected and investigated when discrepancies are observed.

As an overall measure of global accuracy of the delivered radiotherapy doses in the

department, considering more than 5000 individual entrance dose measurements, the mean

dose ratio (measured to expected) is close to unity (1.001) and the standard deviation is

close to 3%. This, of course, also inherently includes the uncertainties associated with the

diode measurements.

Tolerance levels were chosen on the basis of the pilot studies, at approximately 2 SD, with

the aim of not being too wide that significant problems were missed, but not too narrow

that time-consuming investigations were triggered which were inconclusive or which gave

rise to reduced confidence in the system. 5% was selected for entrance dose measurements

and 8% for exit doses (although tighter tolerances of 3% and 6% may be applied for

conformal treatments, especially if dose escalation is involved. However this requires a

greater effort to achieve than for the general routine use).

5.3.4 ROUTINE USE

Having carried out the detailed studies outlined above, multidisciplinary discussion then

centred on how the department was to utilise diode verification dosimetry in routine

practice. For this the daily positioning and recording was to be passed from the research

radiographer to the normal treatment unit radiographers, unit by unit as the routine use was

123

rolled out to the linacs one-by-one. Possible usage and the procedures involved were

evaluated in terms of cost-effectiveness and with the aim of minimising the time involved at

the treatment units, as the patient workload per linac in the department is high.

The decisions on how and wh en to use diodes routinely, and the rationale for each, were:

•

to measure only entrance doses for routine radiotherapy treatments, as the

department’s main aim for routine use was to identify significant errors in treatment,

which had not been picked up by the other levels of the quality system, which includes

independent plan calculation checks, independent MU checks, independent check of

information into the verification system and independent radiographer checks on

treatment parameters and patient set-up. It was felt that entrance dose checks were the

most cost-effective way to do this, independently checking the combination of MU,

beam parameters, beam modifiers, patient position and machine performance in a

relatively simple way. To carry out exit dose checks on a routine basis, there is a

significant increase in the time and resources required, without a similar significant pay -

back.

•

to initially aim to check all patients, building up to this gradually, linac-by-linac, as the

pilot studies had not shown any situations which were significantly worse or better

than others. The aim is to re -assess this decision periodically, taking into account the

outcome information of the diode programme as it progresses.

•

to normally take just one measurement on each patient, which must be carried out

within the first few days of treatment, ideally within the first two fractions, but in no

circumstances more than one week into treatment, so that any problems are identified

early in treatment and rectified. Frequently, dio de measurements are carried out on each

linac on one particular day and all new patients on that machine are monitored. This

implies that the number of patients monitored on each machine per week is generally in

the range of 8 - 12.

124

•

to apply a general tole rance level of 5 % to individual entrance dose measurements, as

discussed above. The action level is also made equal to the tolerance level, i.e. all

deviations above 5 % are investigated.

•

the actions taken if a deviation is observed over this level have evolved with time.

However this department has come independently to a very similar scheme to that

developed in the Leuven department, as discussed in Section 2.2, Figure 2.2 of this

booklet. On-the-spot checks are carried out on the treatment machine when significant

discrepancy is noted. These checks include patient position, diode position, beam

parameters, etc. In addition, at the earliest opportunity, the treatment unit staff notify a

responsible me mber of the physics department, who checks treatment plan,

calculations, treatment information transfer, information in the verification system,

expected signal from the diode, etc. and looks for possible reasons why there may be a

deviation. Whether or not a reason is identified, and unless a trivial reason is

recognised, a second check would be made on the following fraction, with a member of

the physics department present. If this measurement is within tolerance, then the

treatment is deemed acceptable. If not and the same immediate or afterwards checks do

not identify a valid reason, then the physics department organises a phantom study to

simulate the treatment, comparing ion chamber measurements to the diode to decide

whether the treatment should continue without change or not, i.e. investigating both

the clinical irradiation and the diode behaviour in this situation.

In addition to the above, the routine in vivo dose measurement programme aims to carry

out full entrance and exit

dose studies (and estimation of target volume dose from these)

on selected groups of patients for:

•

newly commissioned treatment units, and/or new treatment techniques, or following

major changes in planning systems or planning calculations, to ensure that the whole

system is tested in these circumstances; in particular in case any potential problems

have been overlooked

125

•

full studies on critical patient/treatment groups, for instance dose escalation groups

(with improved tolerances, as discussed above), TBI (but here using TLD), etc.

•

and occasional full audit studies on selected groups of limited numbers of patients, as a

repeated overall check on both the diode methods and on the total radiotherapy

process.

All these essentially mirror the clinical pilot studies in operation and are carried out by

physics personnel and/or the research radiographer, in conjunction with the radiographers

on the particular treatment unit.

5.3.5 METHODS TO SIMPLIFY ROUTINE USE

A number of things have been investigated and some implemented in an attempt to

simplify routine diode use, to make it as cost-effective as possible and in particular to

minimise the time involved at the treatment unit. This has included:

5.3.5.1 POSSIBLE OMISSION OF CORRECTION FACTORS

A full set of correction factors was applied to all the measurements in the clinical pilot

studies. At this stage it is necessary to obtain the best accuracy possible. However for

routine use on routine treatments (although not necessarily so for critical groups, such as

dose escalated patients), it was thought possible that a simpler approach, omitting

correction factors might be applicable. The advantages would include a simpler

methodology and less quality control on the diodes. Initially a comparison was carried out

for patient groups from the clinical pilot studies, where dose estimates made by removing

the correction factors were evaluated. Typically the mean value changed by acceptable

amounts, depending on site, but the standard deviation generally increased significantly.

This implied that an increased tolerance level would have to be used and would have lead

to situations where some significant clinical discrepancies in dose would have not been

recognised. Therefore this approach was deemed not acceptable. Given that the same

126

diode systems are to be used for the more critical groups, it is still necessary to measure

correction factors and to make quality control checks on their values.

5.3.5.2 THE USE OF BUILD-UP CAPS

Build-up caps have been constructed for the diodes to match the build -up more closely to

the beams in the department, with the aim of reducing the spread of correction factors

dependent on secondary electron and photon spectrum effects (see also Section 5.2 for

‘high energy’ beams). Unmodified EDP-10 diodes were appropriate for the 4 MV beam and

unmodified EDP-20 diodes were appropriate for the 8 - 9 MV beams. However, additional

caps of 0.6 mm of brass, in combination with EDP-10 diodes, have been investigated for the

6 MV beams. For the 15 and 16 MV beams, additional caps of 1.2 mm of brass, copper and

stainless steel, in combination with EDP-20 diodes, have been studied. These additional

caps bring the build -up thickness to 15 mm water-equivalent for 6 MV beams and 30 mm

water-equivalent for 15 – 16 MV beams. Calibration and correction factors were measured

for both ‘old -style’ and ‘new-style’ EDP-10 and EDP-20 diodes with these caps. Detailed

results are reported elsewhere ([Blyth 2001a], [Blyth 2001b]).

In summary:

•

the caps produce a reduced range of those entrance correction factors which depend at

least in part on secondary electron and photon spectrum effects, such as field size, tray,

block, wedge, etc.

•

in general, the caps improve the situation to the stage that the correction factors can be

ignored, i.e. the spread is within ± 0.5 %, for instance field size, trays, etc., or a varying

factor can be replaced by one single factor, e.g. for the motorised wedge for all field

sizes

•

the changes were observed to be significantly less for Scanditronix ‘new-style’ diodes

than for ‘old -style’ diodes, in that the newer versions have less variation on some of

these factors to begin with.

•

there is little change in the range of factors for exit measurements

127

•

the range of values for some correction factors, such as for angle of in cidence, are made

worse

•

the shadowing effect is, of course, increased by the use of caps. However as the

number of times that measurements are carried out on an individual patient in the

department is small, even in the full studies where measurements are repeated once per

week, this does not present a significant problem.

•

the gains were less obvious for the high energy beams (15 and 16 MV) than for the 6

MV beams.

Additional build -up caps of 0.6 mm brass are currently routinely used for all our 6 MV beam

diode measurements. The routine use of build -up caps for 15 and 16 MV beams is still

under discussion.

5.3.5.3 THE USE OF ‘GENERIC’ CORRECTION FACTORS

The use of ‘generic’ correction factors has been investigated for the case of standard

diodes as well as for dio des with build -up caps. For a given treatment modality and for a

specific treatment field, the range of treatment parameters has been investigated for a

representative sample of patients. For each relevant parameter the range of the appropriate

correction factor has been considered and a combined correction factor calculated from the

mid-range values (or the values judged to be most representative). In some cases this has

required a judgement on, for example, the range of beam fractions that are wedged and

unwedged in particular clinical situations (for motorised wedge machines) and a weighted

wedged/unwedged correction factor to be included, etc. In general, the range of overall

correction factors around this generic correction factor is small for the norma l range of

treatment parameters used for different patients for a particular field in a particular type of

treatment. Also this is generally better when build -up caps are used, as the range of a

number of correction factors is reduced.

128

Generic correction factors are currently in use for routine entrance dose measurements for

our 6 MV beams using diodes with build -up caps. Tables of generic correction factors are

available, listed by diode, treatment unit and modality (where normally one specific diode

and electrometer combination are assigned routinely to a given treatment modality), by

treatment technique and by treatment field. In this way, only one factor is required and is

easily available in any particular situation. This approach implies that if an ou t-of-tolerance

value is observed, one of the things that the investigating physicist does first is check that

the particular treatment parameters used for that patient are within a tolerable range of the

factors used to produce the generic factor. In no case so far has the use of generic factors,

coupled with build -up caps, given rise to a problem.

Full correction factors are still used in any audit studies, critical group studies and new

equipment or new technique studies.

The application of generic correction factors is also in use in Amsterdam [Meijer 2001]. The

time required for analyzing the patient measurements is hereby substantially reduced, while

keeping the accuracy within acceptable limits. For prostate treatments, the additional

uncertainty for the target absorbed dose as a consequence of the interpatient varience of

the diode correction factors is estimated to be 0.2 % (1 SD). A similar, though more

radically simplifying approach is used by Alecu et al. [Alecu 1998]. They eliminate the

necessity of measuring seperate diode correction factors by using a second calibration

factor, having as “reference conditions” the average conditions of specific routine

treatment situations that deviate the most from the usual reference calibration conditions.

5.3.5.4 DATA COMMUNICATION AND RECORDING

One of the aims of the department in considering routine diode implementation was to

minimise the time necessary at the treatment unit and also to minimise the duplication of

effort. Therefore for the routine use of diodes fo r monitoring standard treatments, some

consideration was given to the calculation of expected measurement values and to how the

radiographers were required to record results.

129

Throughout the initial and follow-up studies, all measurements have been recorded

manually onto separate in vivo dosimetry sheets and then entered into a spreadsheet by

the research radiographer (or other research student involved in the project). The

calibration and correction factors are applied in the spreadsheet and the resulting dose

compared to the expected value, also calculated there from data input from the

prescription/treatment sheet and from planning data for the patient and for the treatment

machine.

For routine use, the cost-effective method adopted for calculating the exp ected result was

for physics staff, at the treatment planning stage, to produce an expected diode entrance

reading, taking the expected daily given dose and dividing this by the calibration factor for

the diode/treatment modality and by the generic correction factor for the diode, treatment

modality, treatment technique and field. This is easily done at the time of planning and MU

calculation when the given dose is being recorded. At the same time the ± 5 % tolerance is

applied to this value, so that a range of readings is written on the treatment sheet in an in

vivo dosimetry section. The treatment unit radiographers then simply have to check that

the measured reading is within this range and, at the simplest level, tick one box. If the

measurement is not within range, they must cross another box and place the sheet in a tray

for reference to the physics group for further investigation. In practice this process is

simple to operate. For example all breast patients on any particular treatment unit have the

same range of required diode readings, provided the dose and number of fractions is

standard. For our matched units from the same manufacturer, this same range applies for all

breast patients on both units, etc.

One consequence is that this does not record numerical data. However, if required for

analysis, this is obtainable either by requesting the radiographers to write down the

reading, as well as ticking one or other of the boxes, or by directly grabbing the readings

via the electrometer/PC interface. In general, we have not done either of these things for

routine data, although we may occasionally do so for audits. We intend to regularly

quantitatively assess the performance of the systems by limited patient number studies

130

repeating full entrance and exit measurements, as discussed above, and currently feel there

is no pressing need to analyse additional data.

5.3.5.5 DIODE MOUNTING AND HANDLING

In the initial studies, the diodes were connected via cables following the normal dosimetry

channel route from the tre atment room out to the control area. Following this in the early

routine implementation, permanent under-floor cables were installed to remove as much

cable from the floor area as possible and permanent junction boxes were installed on the

wall in the treatment room. Diode mounts on the couch and on the gantry were considered

and some prototype designs studied. Whilst the dedicated research radiographer, or other

research students, were involved in carrying out the measurements this was not a problem,

as they took care of the cabling, the position of the diode throughout patient setup and the

positioning of the diode. This was independent of what the treatment unit radiographers

were doing and practical and logistics problems were minimal. A number of instan ces of

connector damage were noted due to excess strain on the connectors from the hanging

cables, or due to damage when connectors were on the floor or were trapped between the

floor and the rotating patient support system. When the measurements were beco ming

routine and were being rolled out to the treatment unit radiographers, this system was not

acceptable. Instead a simple rotating mount was installed on the ceiling above the

treatment unit, directly above and approximately half-way along the horizontal arm of the

rotating gantry. From this an inverted L-shaped cable support was suspended, made from

light-weight cylindrical pipe, with the cable down the centre. The cable goes through the

centre of the rotating mount and above the false ceiling to a cable-way out of the room to

the control area where the electrometer is sited. At the other end it terminates in a

connector at the end of the pipe. The diode connects to this and rests on a quick-remove

hook at the base of the pipe. The swinging cable support is very easily swung completely

out of the way for patient setup and in towards the isocentre for diode positioning. The

system is shimmed to hold position at any point. The height of the lower edge of the cable -

132

5.4 LARGE SCALE IN VIVO DOSIMETRY IMPLEMENTATION – THE

COPENHAGEN EXPERIENC E

5.4.1 INTRODUCTION

In 1999 in vivo diode dosimetry was implemented in the Finsen Centre (FC) in Copenhagen,

Denmark. We started with a small group of FC patients, selected by virtue of their relatively

simple treatments. The initial purpose was to develop an effective and reliable quality

assurance procedure and subsequently to include treatments of greater complexity. This

has gradually been achieved. In the FC approximately 2250 patients are treated per year.

Seven linear accelerators (linacs) are available with energies ranging from 4 MV to 18 MV.

At this stage diode measurements are carried out on all linacs but there are still specific

types of treatment to be incorporated into the procedure. Our intention was to perform the

procedural measurements with every FC patient. In Denmark the legislation demands a

protocol where every patient treated should be undergoing in vivo dosimetry commencing

at the start of the treatment course. Taking this into consideration and the capacity of the

centre, our approach has been to utilise an in vivo dosimetry (IVD) system with a relatively

broad tolerance window aiming at detecting large deviations.

5.4.2 METHODOLOGY

The patient dose measurement is carried out in the beginning of the treatment course,

within the first three fractions in order to make it possible to correct any detected errors.

Expected diode values are derived from an independent spread sheet program containing a

database of correction factors and beam data (depth dose distributions). The program is

not integrated with the Record and Verify system (R&V), hence the prescribed dose and

beam parameters are manually entered into the spreadsheet. Clearly this procedure

increases the number of errors in the quality control (QC) process (errors in input data) but

constitutes an independent calculation of the expected entrance dose. However, in case of

133

computerised planning, the entrance dose at a point on the central axis c alculated with the

TPS is entered and relevant correction factors applied.

As asymmetric field technique (half- or three quarter of the field blocked) is standard

practice at the FC, the position of the diode during measurements has to be determined

during the IVD preparation. Therefore additional corrections of the expected diode reading

at the central axis are added in the spreadsheet to account for various focus -to-diode

distances and off-axis positions in wedged fields and treatment fields modulated wit h a

compensating filter. In a situation where the diode is positioned on immobilisation devices,

couch or bolus, the actual focus-to-diode distance is recorded during treatment, followed

by a correction of the expected diode reading.

5.4.3 EQUIPMENT

We sought a QC process with great simplicity and cost-benefit advantage. Therefore we

chose electrometers easy to handle and diodes with low sensitivity degradation with

accumulated dose, in order to reduce the need for repetitious calibration. Presently we use

the Apollo-5 electrometers and diodes P10 (4 MV), P20 (6-8 MV) and P30 (18 MV) (MDS

Nordion AB). In the tangential treatment technique, measurements are carried out with

cylindrically shaped diodes, Isorad-p diodes (Sun Nuclear Corporation), to minimise the

influence of diode directional dependence thus reducing the number of correction factors.

One treatment unit is equipped with QED diodes (Sun Nuclear Corporation). The use of

different types of diodes will allow us to evaluate statistical fluctuations due to various

diode characteristics.

5.4.4 CALIBRATION PROCEDURE

Currently, the calibration frequency is once every third month. Typically, the sensitivity

has decreased by less then 0.5 % during this period of time (corresponding to around

400Gy). The calibration is performed with a Solid Water

phantom in conjunction with the

weekly constancy dosimetry check (ion chamber in plastic phantom) of the treatment unit.

134

Here, some aspects have to be considered regarding the type of errors one aims to detect.

The diode s ystem may be calibrated against:

1) an absolute dosimetry system

2) a constancy dosimetry check system

3) the monitor chamber of the linear accelerator

The ultimate choice of calibration method is to calibrate the IVD system to absorbed dose-

to-water using an ionisation chamber with a traceable calibration factor. Any malfunctions

of the treatment unit as well as human errors, such as an erroneous calibration of the

treatment unit, are then detected. However, this calibration procedure is onerous and time -

consuming. If the QC system is calibrated against a constancy check of the linac output

(accelerator weekly output check), the second suggested calibration method, a

malfunctioning of the treatment unit is most likely to be detected while an erroneous unit

calibration may be veiled. The third, commonly used method would probably not detect

deviations related to an erroneous calibration of the treatment unit. Furthermore, a

systematic error may be introduced if the diode QC system is calibrated against the mo nitor

chamber of the linac i.e. the diodes are adjusted to the level of that linac’s specific day’s

output. Eventhough modern linear accelerators have a high stability (constancy within

%) this may be of importance in studies where a high accuracy is required in the QC

process (when diode measurements are used to check dose delivery to the target volume).

However, the uncertainty in the diode dose determination is a combination of many

parameters and it is not likely that any day-to-day variations in linac performance in

individual patient measurements will be detected.

Considering the type of deviations that we aimed to detect, we believed that diode

calibration using a constancy dosimetry check system was most cost-effective.

The variation in diode re sponse with temperature was accounted for in the calibration

procedure by means of adding a temperature correction factor to the diode reading during

135

calibration. However, it was considered inappropriate to include the influence of

temperature dependence in the head & neck region where immobilisation devices are used

in most cases.

5.4.5 CORRECTION FACTORS

This section does not deal with diode characteristics or with the variation of different

correction factors, as they are presented and discussed in Chapter 1 and Sections 5.1 and

5.2. Correction factors were applied to account for field size dependence and variations in

response at different SSDs in the energy range of 8 – 18 MV. No wedge correction factor

was necessary as only dynamic wedges are used in the FC. However, a correction factor

accounting for the non-linear dose per pulse dependence of the n -type diodes was applied

4 MV treatment fields modulated with compensation filters

Temperature dependence was

accounted for in the calibration procedure. In tangential treatment fields the cylindrically

shaped diode has been adopted, consequently no correction of the directional dependence

was required.

In order to reduce the number of correction factors , each diode type (P10, P20) had one set

of correction factors per energy; i.e. the same factors were applied regardless of treatment

unit or the individual diode.

5.4.6 TOLERANCE LEVELS

Fairly broad tolerance levels were chosen during the initial phase of the implementation

with the intention that they would be gradually minimised with hindsight. Reasonable

levels were established from phantom measurements and selected patient measurements.

The tolerance levels, coinciding with action levels, were related to the complexity of the

treatment delivery according to Table 5.9, rather than the intention of the treatment

(radical/palliative).

136

Treatment

site

Treatment technique

Tolerance

level

Diode type

Breast,

lumpectomy

Tangential field s, 6 - 8 MV

½ field blocked, dynamic wedge

Isorad-p (Sun Nuclear)

Cylindrical build-up cap

Breast,

mastectomy

Anterior field, 6 - 18 MV

¾ field blocked

P20, P30

(MDS Nordion AB)

QED 1115/1116

(Sun Nuclear)

Head &

Neck

Patient individual field t echnique

4 – 6 MV, ½ field blocked

compensating filter/dynamic

wedge

P10, P20

(MDS Nordion AB)

QED 1115 (Sun Nuclear)

Chest/Pelvic Patient individual field technique

6 – 18 MV

P20, P30

(MDS Nordion AB)

QED 1115/1116

(Sun Nuclear)

Table 5.9 Tolerance and action levels established for different treatment sites and the

diode type used.

5.4.7 RESULTS AND DISCUSSION

Since the start in 1999, over 3000 treatment fields have been monitored by means of diode

measurements of entrance dose. Figure 5.16, Figure 5.17 and Figure 5.18 shows of a major

part of all measurements as a percentage deviations from expected values, grouped in bins

of 1 %.

137

Figure 5.16 Deviations from expected value of diode measurements in manually planned

treatments. Breast treatments excluded.

Figure 5.17 Deviations from expected value of diode measurements in computer planned

treatments. Breast treatments excluded.

<-9.5

-9

-8

-7

-6

-5

-3

-2

-1

> 9.5

Number of fields

% Deviation from expected value

4 MV

6 – 8 MV

18 MV

4 MV

6 - 8 MV

18 MV

Average

2.64

0.43

0.67

1 SD

2.9

2.3

2.2

Number

173

347

392

100

120

140

160

% Deviation from expected value

Number of fields

4 MV

6 – 8 MV

18 MV

4 MV

6 - 8 MV

18 MV

Average

1.64

0.31

1.2

1 SD

3.0

2.7

2.2

Number

776

329

699

<-9.5

-9

-8

-7

-6

-5

-3

-2

-1

> 9.5

138

Figure 5.18 Deviations from expected value of diode measurements in breast treatments

(computer planned treatments).

The positioning of the diode becomes more critical in treatment fields incorporating

wedges or compensating filters, reducing the precision in the QC process. This can also be

seen as a larger deviation in the frequency distribution. In simple, manually planned

treatments the level of complexity is low (no wedges, compensation filters or field

asymmetry). Consequently, the variations from the expected value are smaller; i.e. for

energies 6 – 8 MV a standard deviation of 2.3 % (1 SD) was calculated and 2.2 % for 18 MV

(Figure 5.16). In more sophisticated, computer planned treatments (Figure 5.17) the

corresponding figures were slightly higher for 6 - 8 MV: 2.7 %. The largest spread was

recorded in the tangential treatment technique with a standard deviation of nearly 5 %.

This would, to some degree, be expected because of the higher level of complexity. The

measurements of mastectomy treatment patients showed a significant shift towards

positive variation suggesting a systematic error. The distribution had an average of 3.5 %

with a standard deviation of 3.1 %. An investigation of possible causes revealed errors in

Number of fields

Mastectomy

Lumpectomy

<-9.5

-9

-8

-7

-6

-5

-3

-2

-1

> 9.5

% Deviation from expected value

Mastectomy

Lumpectomy

Average

3.51

-1.03

1 SD

3.12

4.71

Number

289

149

139

the output factors used to calculate monitor units in quarter fields (¾ blocked fi eld) i.e. the

shift of the average was induced by systematic errors in the treatment process.

At the end of the year 2000, one of our 4 MV energy units had a breakdown and during the

subsequent absolute dose-to-water calibration the output was, by human error, adjusted

nearly 5% higher than intented. The diode system was not recalibrated at the same time

and consequently, the erroneous calibration was detected with patient diode

measurements. However, these measurements (a total of 27) on their own, do not explain

the shift in the mean average of the distribution in 4 MV energy treatment fields. The shift

shows an average of 1.6 % in computerised planned treatments and 2.6 % in manually

planned treatments. This deviation is believed to indicate a systematic error in the QC

process rather than the treatment process, considering the fact that the dosimetry accuracy

in this treatment technique has been carefully verified. The shift has to be further

investigated. One aspect may be reconsidered: the diode temperature dependence that is

not accounted for in 4 MV treatment fields. Further actions have to be taken to improve the

accuracy in the diode positioning in head, neck and breast treatments.

Besides the erroneous unit calibration and the systematic errors in the dose calculations

the diode QC process so far has detected two errors in the treatment process: 1) A

discrepancy in beam energy between calculation (manually planned) and treatment delivery

caused by an erroneous input in the R&V system. 2) For a patie nt with two target volumes,

the treatment fields were mixed up; caudal target volume was treated with cranial treatment

fields and vice versa.

5.4.8 CONCLUSION

The implementation of a new QC tool, such as diode measurements of entrance doses, is an

ongoing process that continuously needs evaluation and refinement. Our goal was to

implement diode measurements for every patient undergoing external radiotherapy in the

FC. To achieve an acceptable workload, a broad tolerance window was defined, accepting

relatively large deviations.

141

5.5 RESULTS OF SYSTEMATIC IN VIVO ENTRANCE DOSIMETRY – THE

MILANO (HSR) EXPERIENCE

Systematic entrance dose in vivo dosimetry was gradually implemented at our Institute

from the end of ’94 in one of the three treatment rooms. It has recently been extended to

another treatment room (November 1999), and the extension to the third room is currently in

progress (Spring 2001).

In this review we present our experience during the period November ’94 – April 2000 on

more than 3900 measurements referring to 2001 patients.

5.5.1 MATERIALS AND METHODS

5.5.1.1 EQUIPMENT

The two treatment rooms where in vivo dosimetry was implemented are provided with a 6

MV linear accelerator (Linac 6/100); no record and verify systems are available at these

facilities.

For the entrance dose measurements we used p-type silicon diodes (Scanditronix EDP 10)

connected to a multi-channel electrometer (DPD 510, DPD 3 Scanditronix). The diodes are

monthly calibrated against an ionisation chamber in a reference geometry to convert the

diode signal in entrance absorbed dose. Procedures for the calibration are similar to those

reported in the ESTRO Booklet n° 1 [Van Dam 1994]. Cadplan (Varian-Dosetek Oy, versions

2.6.2, 2.7, 3.1) is our 3D treatment planning system (TPS); however, during the period

November ’94 - June ’95 we used (2D) older versions of Cadplan (2.5, 2.6.1).

5.5.1.2 IN VIVO MEASURED AND EXPECTED ENTRANCE DOSE

In vivo entrance dosimetry was generally performed at the first session of the therapy

(always within the first three or four days from the start of the treatment). After patient set -

up, just before the start of the irradiation, the diode was positioned at the centre of the

irradiation field, except in the case where the centre is shielded by blocks: in this case the

142

diode was positioned far from the penumbra region. Before fixing the diode, the source -to-

skin distance (SSD) was read and compared with the value defined during the simulator

session and used for treatment planning. In this way possible effects on the prescribed

dose due to small discrepancies between the two conditions (planning/therapy) could be

assessed during interpretation of the in vivo measurement.

Diode readings were not corrected by inverse square correction factors due to the choice

of monitoring the “true” accuracy of treatment delivery. Only if the focus-to-diode distance

was different from the SSD because of the use of immobilisation systems or other ancillary

equipment, the diode reading was corrected by an appropriate inverse square correction

factor.

The measured entrance dose was defined as the diode reading corrected by correction

factors. The influence on the diode signal of collimator opening, source-skin distance

(SSD), wedges, tray and obliquit y of the beam with respect to the diode axis was

previously investigated and appropriate correction factors were determined. The diode

signals were initially not corrected to take temperature into account. Only during the last

year we introduced a “clinical” correction factor for temperature. Due to the limited time

when diodes are in contact with the skin in clinical conditions (around 1 – 2 min.), based on

our measurements, it was estimated to be small (less than 0.5 % correction).

The expected entrance dose was calculated from the prescribed tumour dose by an

independent formula, based on tabulated TPR values and peak scatter factors for the

appropriate equivalent field sizes.

5.5.1.3 QA CHAIN: METHODS

A 5% action level was applied. However, in the case of opposed beams, a larger error was

accepted in one of the two beams (up to 7-8 %), if the average percentage deviation of the

two opposed fields was within the 5% limit. In this way, we took into account that small

SSD errors may compensate each other in terms o f isocentre dose. The 5% action level was

chosen on the basis of the expected accuracy and reproducibility of our measurement

143

system. This level roughly corresponds to 2 SD where SD is the global standard deviation

of the entrance dose measurement accuracy in most critical conditions (blocked and

wedged fields, tangential beams), excluding patient set-up inaccuracy. The global standard

deviation includes the accuracy and the reproducibility of the entrance dose measurements

with the ionisation chamber and with the diode system and the single SDs of the measured

correction factors.

If the discrepancy between the measured entrance dose and the expected one was below

the action level, in vivo dosimetry data (measured dose, expected dose and the percentage

deviation) was registered by a physicist with the appropriate of information for further

statistical analysis.

When the action level was exceeded, all treatment parameters were verified and the

measurement was generally repeated; a physicist directly performed all these phases (the

repetition of the in vivo dosimetry check may be performed by the technician, often with

the presence of a physicist). Firstly, the agreement between the treatment planning data

and the corresponding simulator data with the control of the data transfer on the treatment

chart was carefully checked. MU calculation/data transfer errors were expected to be rare

because a double check of the treatment chart data, of the treatment planning and of the

MU calculation was always done before tre atment delivery. Further checks concerned the

right use of wedges and blocks, including the presence of a block near the irradiation field

centre. If one of these checks was positive, the percentage deviation between the entrance

measured dose and the expected one could be explained. After correcting the mistake, a

second in vivo dosimetry check was performed to confirm the agreement between the

measured and the expected dose.

These checks were not always able to explain the deviation between the measured e ntrance

in vivo dose and the calculated one. Irrespective of this, a second check was always

performed. If a “large” deviation was detected even after the second check, we sometimes

measured the entrance dose on a solid phantom with diode and ionisation ch amber in the

same treatment conditions (i.e.: field width, wedge, blocks, SSDs…) and compared it with

144

the expected one. Because of the accumulated experience in assessing the causes of the

persistent deviations, the number of phantom controls decreased wit h time.

5.5.1.4 MU CALCULATION/DATA TRANSFER CHECK

From 1991 a double check procedure of MU calculation and data transfer/treatment chart

compilation has been implemented in our department. In previous reports we demonstrated

the ability of this simple tool in strongly reducing the occurrence of systematic errors

before treatment delivery. Every day a “controller” checks the MU and the dose

distribution calculation that has been performed by another operator, together with a check

of the irradiation data reported o n the treatment chart.

5.5.2 RESULTS

5.5.2.1 DETECTION OF SYSTEMATIC ERRORS

The in vivo dose measurements on 2001 patients revealed 14 systematic errors. 12 (0.6 %)

were serious (i.e.: leading to an under/over-dosage larger than 5 %) and 6 (0.3 %) were

larger than or equal to 10 %. In Table 5.10 the causes of all errors are reported. If excluding

“thickness” errors which cannot be detected by the MU calculation/data transfer check,

the rate of serious errors detected by in vivo dosimetry (which “really” escape the MU

calculation/data transfer check) was equal to 0.4 %. A number of minor errors due to an

uncorrected assessment of patient thickness were not considered; a number of random

errors (i.e. occurring for one fraction) we re also not considered.

In vivo dosimetry also permitted us to promptly detect a systematic error in measuring

patient thickness occurring during ‘97, due to a bad resetting of the “zero” indicator of the

simulator couch.

145

5.5.2.2 SYSTEMATIC ERRORS DETECTED BEFORE IN VIVO DOSIMETRY BY MU

CALCULATION/DATA TRA NSFER CHECK

In Table 5.10 the systematic errors detected before delivering the treatment by a check of

the treatment chart, of the treatment planning and of the MU calculation are reported. Data

refer to a longer period (‘91-‘99) and to patients treated in all treatment rooms. The rate of

serious errors was found to be equal to 1.53 % with a 0.77 % rate for errors larger than 10

5.5.2.3 PATIENTS WITH MORE THAN ONE CHECK

Globally, 156/2001 (7.8 %) patients underwent more than one check; only 6/2001 had more

than 2 checks (see Table 5.11). The rate of second checks was higher for breast patients

(10.2 %) against non-breast patients (7.3 %, p = 0.06).

5.5.2.4 ACCURACY OF TREATMENT DELIVERY

In Table 5.12 the values of mean, median and SD of the distribution of the deviations

between measured and expected doses are reported together with the rate of deviations

larger than 5, 7 and 10 %. Globally, we had a mean deviation equal to 0.2 % with a SD equal

to 3.1 %. The rates of deviations larger than 5, 7 and 10 % resulted to be 10.3, 2.6 and 0.2 %

respectively. When averaging the deviations of opposed beams, the standard deviation

was 2.7 % and the rates of deviations larger than 5, 7 and 10 % reduced to 4.9, 0.8 and 0.0 %

respectively.

When considering the presence of a wedge, the rate of deviations larger than 5 and 7 %

was significantly higher than in the group without wedge (p < 0.0001). Similarly, the

presence of a block was related to a higher rate of deviations larger than 5 and 7 %

(respectively p < 0.03) and a systematic deviation from 0 (+ 1.5, p < 0.001).

147

In vivo dosimetry

MU calculation and

data transfer check

Period

NOV. ‘94 – APRIL 2000

Sept. ’91 – May ‘99

Beams

Photons 6 MV

Photons 6, 18 MV

Electrons 6, 9, 12, 16 MeV

n° checks

3932

9747

n° patients

2001

7700*

Errors (including minor errors)°

14**

362

Errors > 5 %°

118

Error rate =

(n° errors > 5%)/n° patients

0.60 %

1.53 %

Type of errors (> 5 %) and rate:

Data

Energy

Blocks/Equivalent field

TPS (bad data entry)

Normalisation

Wedge

(all excluding thickness)

Thickness*** °

1 (0.05 %)

2 (0.10 %)

3 (0.15 %)

2 (0.10 %)

(8 (0.40 %))

4 (0.20 %)

42 (0.55 %)

14 (0.18 %)

8 (0.10 %)

3 (0.03 %)

41 (0.53 %)

10 (0.13 %)

Errors

≥

10 % (error rate)

6 (0.30 %)

59 (0.77 %)

* Estimated

** Minor errors due to uncorrect assessment of thickness were not considered

*** Large errors due to wrong assessment of thickness

° Excluding 5 large “thickness” errors due to bad resetting of simulator couch(see

text)

Table 5.10 The systematic errors detected by in vivo dosimetry are shown together

with those detected by MU calculation and data transfer check. Serious

148

systematic errors were defined as those which, if undetected, could lead to a

5 % or more error on the delivered dose to the PTV.

N pts

Patients with more than one check

156

2001

7.8

Patients with more than two checks

2001

0.3

Breast

362

10.2

Mediastinum-abdomen – pelvis

(AP-PA fields)

640

7.3

Neck (lateral fields)

179

7.3

Arms -limbs

103

8.7

Brain

307

7.5

Others

410

5.6

Table 5.11 The rates of repetition of the in vivo dosimetry check.

Beams

MEAN Median

%>5% %>7% %>10%

All

3770

0.2

3.1

10.3

2.6

0.2

All (averaging opposed beams)

2095

0.2

2.7

4.9

0.8

0.0

1055

-0.2

2.6

5.5

1.4

0.0

958

1.0

0.8

3.2

12.7

3.3

0.2

Lateral

957

-0.2

3.0

8.3

1.8

0.2

Oblique

799

0.4

0.7

3.5

15.6

4.5

0.4

Wedged

954

0.6

0.8

3.5

15.6

4.8

0.4

Unwedged

2816

0.1

0.0

2.9

8.3

1.9

0.1

Blocked

1259

1.5

3.0

11.8

3.5

0.4

Blocked & unwedged

1156

1.4

1.5

2.9

10.7

2.8

0.3

149

Unblocked

2511

-0.4

-0.5

3.0

9.4

2.2

0.1

Unblocked & unwedged

1660

-0.8

-0.9

2.6

6.7

1.3

0.0

Breast

719

0.3

0.5

3.5

15.7

3.9

0.3

Brain

593

-1.0

-1.2

2.8

8.4

1.3

0.0

Neck (lateral)

344

1.1

1.4

2.8

8.1

2.3

0.0

Neck (AP-PA) &

Supraclavicular

103

1.0

3.3

10.7

5.8

1.0

pelvis, abdomen, thorax AP/PA

1261

0.5

0.3

3.0

10.1

2.0

0.0

ARMS/LIMBS

160

0.1

-0.2

3.2

10.6

6.3

0.0

VERTEBRAE

280

0.1

0.3

2.1

1.0

0.5

0.0

Table 5.12 Deviations between measured and expected entrance dose: mean and median

deviations, SD, rates of deviations larger than 5, 7 and 10 % are presented in

a number of ways. If more than one check was performed, the data of the last

check were considered (n = 3932, 2001 patients).

5.5.3 FINAL REMARKS

Our experience in systematic in vivo dosimetry confirms that a number of serious

systematic errors might escape the independent check of dose calculation and data

transfer, which should be always performed before treatment delivery. Moreover, in vivo

dosimetry permits detection of a number of minor errors (SSDs and thickness errors) which

would be undetected by the independent check, thus improving the global quality of the

treatment. The high accuracy which can be reached by in vivo dosimetry with diodes, once

appropriate correction factors are applied, can also, by pooling patient data, detect

machine-related problems (for example, the discovered bad resetting of the simulator

couch), wrong configuration of treatment units on TPS and uncorrected procedures during

the chain which precedes treatment delivery, wh ich could lead to systematic errors in dose

150

delivery on a large number of patients. Another important goal of in vivo dosimetry is

maintaining a high level of attention on quality by all the involved staff. For this reason, in

our opinion, it is reasonable to suppose that without in vivo dosimetry, the rate of serious

systematic errors could be higher than the one reported by ourselves and in other similar

studies. In our opinion, any effort in implementing systematic in vivo dosimetry is justified.

However, it is mandatory to consider that clinical implementation of such programs implies,

above all, the allocation of human resources and a relevant organisational work. This is the

main cause of the limitation of systematic in vivo dosimetry to one of the three treatment

rooms of our institute until November 1999.

A very important element is the need of precise indications if the action level is exceeded.

Too high a number of second in vivo dosimetry checks may induce a negative impression

concerning the aim of the check and could generate distrust among personnel. Our results

indicate that the 5 % choice for the action level was appropriate. However, it could be

useful to set different action levels depending on the type of beam: in our case it will be

reasonable to set a 6-7 % action level for tangential wedged beams.

151

APPENDIX 1: LITERATURE OVERVIEW

[AAPM 1994]

AAPM

Comprehensive QA for radiation oncology

Report of AAPM Radiation Therapy Committee TG40.

Med. Phys. 21: 581-618 (1994)

[

Alecu 1997

]

R. Alecu, J.J. Feldmeier and M. Alecu

Dose perturbations due to in vivo dosimetry with diodes

Radiother. Oncol. 42: 289-291 (1997)

[

Alecu 1998

]

R. Alecu, M. Alecu, and T.G. Ochran

A method to improve the effectiveness of diode in vivo dosimetry

Med. Phys. 25: 746-749 (1998)

[Alecu 1999]

R. Alecu, T. Loomis, J. Alecu and T.G. Ochran

Guidelines on the implementation of diode in vivo dosimetry programs

for photon and electron external beam therapy

Med. Dosim. 24: 5-12 (1999)

[

Aletti 1991

]

P. Aletti, P. Bey, A. Noel, L. Malissard and C. Sobczyk

Programme d’assurance de qualité dans l’exécution de la radiothérapie

externe au Centre Alexis Vautrin

Bull. Cancer/Radiother. 78: 535-541 (1991)

[

Aletti 1994

]

P. Aletti

In vivo dosimetry: measurement strategy in vivo

Bull. Cancer/Radiother. 81: 453-455 (1994)

[

Altshuler 1989

]

C. Altshuler, M. Resbeut, D. Maraninchi, J.P. Guillet, D. Blaise, A.M.

Stoppa and Y. Carcasonne

Fractionated total body irradiation and allogeneic bone marrow

transplantation for standard risk leuke mia

Radiother. Oncol. 16: 289-295 (1989)

152

[

Aukett 1991

]

R.J. Aukett

A comparison of semiconductor and thermoluminescent dosemeters for

in vivo dosimetry

Br. J. of Radiol. 64: 947-952 (1991)

[Bagne 1977]

F. Bagne

A comprehensive study of LiF TL response to high energy photons and

electrons

Radiology 123: 753-760 (1977)

[Bartolotta 1995] A. Bartolotta, M. Brai, V. Caputo, R. Di Liberto, D. Di Mariano, G.

Ferrara, P. Puccio and A. Sansone Santamaria

The response behaviour of LiF:Mg, Cu, P thermoluminescence

dosimeters to high-energy electron beams used in radiotherapy

Phys. Med. Biol. 40: 211-220 (1995)

[Bascuas 1977]

J.L. Bascuas, J. Chavaudra, G. Vauthier and J. Dutreix

Intérêt des mesures "in vivo" systématiques en radiothérapie

J. Radiol. Electrol. 58: 701-708 (1977)

[Beddar 1992a]

A.S. Beddar, T.R. Mackie and F.H. Attix

Water-equivalent plastic scintillation detectors for high-energy beam

dosimetry: I. Physical characteristics and theoretical considerations

Phys. Med. Biol. 37: 1883-1900 (1992)

[Beddar 1992b]

A.S. Beddar, T.R. Mackie and F.H. Attix

Water-equivalent plastic scintillation detectors for high-energy beam

dosimetry: II. Properties and measurements

Phys. Med. Biol. 37: 1901-1913 (1992)

[Biggs 1979]

P.J. Biggs and C.C. Ling

Electrons as the cause of the observed dmax shift with field size in high-

energy photon beams

Med. Phys. 6: 291-295 (1979)

153

[Bjärngard 1993] B.E. Bjärngard, P. Vadash and T. Zhu

Doses near the surface in high energy x-ray beams

Med. Phys. 22: 465-468 (1995)

[Blake 1990]

S.W. Blake, D.E. Bonnett, and J. Finch

A comparison of two methods of in vivo dosimetry for a high energy

neutron beam

Br. J. of Radiol. 63: 476-481 (1990)

[Blanco 1987]

S. Blanco, M.A. Lopez-Bote and M. Desco

Quality Assurance in radiation therapy: systematic evaluation of errors

during the treatment execution

Radiother. Oncol. 8: 256 – 261 (1987)

[Blyth 1997]

Blyth, A.S. Macleod and D.I. Thwaites

A pilot study of the use of in vivo dosimetry for quality assurance in

radiotherapy

Radiography 3: 131-142 (1997)

[Blyth 2001a]

C. Blyth, M. Paolucci, C. Stacey and D.I. Thwaites

Patient and phantom measurements using diodes with and without

buildup caps for dosimetry verification in breast radiotherapy treatment

submitted to Radioth. Oncol.

[Blyth 2001b]

C. Blyth and D.I. Thwaites

Practical cost-effective implementation of routine diode dosimetry for

quality assurance in radiotherapy

submitted to J. Radioth. in Practice

[Boellaard 1998] R. Boellaard

In vivo dosimetry with a liquid -filled electronic portal imaging device

PhD. Thesis, Vrije Universiteit Amsterdam (1998)

154

[Bolla 1995]

M. Bolla, H. Bartelink, G. Garavaglia, D. Gonzalez, J.C. Horiot, K.A.

Johansson, G. van Tienhoven, K. Vantongelen and M. van Glabbeke

EORTC guidelines for writing protocols for clinical trials of radiotherapy

Radiother. Oncol. 36: 1-8 (1995)

[Briot 1990]

E. Briot, A. Dutreix and A. Bridier

Dosimetry for total body irradiation

Radiother. Oncol. Suppl. 1: 16-29 (1990)

[Broerse 1987]

J.J. Broerse and A.E. Varekamp

Total body irradiation for treatment of haematological diseases

Radiother. Oncol. 9: 87-90 (1987)

[Brown 1996]

L. Brown

Development of a diode-based in vivo dosimetry system for clinical use

in radiotherapy

M.Sc. thesis, University of Glasgow (1996)

[

Calandrino 1993

]

R. Calandrino, G.M. Cattaneo, A. Del Vecchio, C. Fiorino, B. Longobardi

and P. Signorotto

Human errors in the calculation of monitor units in clinical radiotherapy

practice

Radiother. Oncol. 28: 86-88 (1993)

[Calandrino 1997] R. Calandrino, G.M. Cattaneo, C. Fiorino, B. Longobardi, P. Mangili and

P. Signorotto

Detection of systematic errors in external radiotherapy before treament

delivery

Radiother. Oncol. 45: 272-274 (1997)

[Carruthers 1999] L. Carruthers, A.T. Redpath and I. Kunkler

The use of compensators to optimise the three dimensional dose

distribution in radiotherapy of intact breast

155

Radioth. Oncol. 50: 291-300 (1999)

[Chavaudra 1976] J. Chavaudra, G. Marinello, A.M. Brulé and J. Nguyen

Utilisation pratique du borate de lithium en dosimétrie par

thermoluminescence

J. Radiol. Electrol. 57: 435-445 (1976)

[Ciocca 1991]

M. Ciocca, L. Landoni, C. Italia, P. Montanaro, P. Canesi and R. Valdagni

Quality control in the conservative treatment of breast cancer: patient

dosimetry using silicon detectors

Radiother. Oncol. 22: 304-307 (1991)

[Contento 1984] G. Contento, M.R. Malisan and R. Padovani

Response of thermoluminescence dosemeters to beta radiation and skin

dose assessment

Phys. Med. Biol. 29: 661-678 (1984)

[Cozzi 1998]

L. Cozzi and A. Fogliata-Cozzi

Quality assurance in radiation oncology. A study of feasibility and

impact on action levels of an in vivo dosimetry program during breast

cancer irradiation

Radiother. Oncol. 47: 29-36 (1998)

[Cross 1992]

P. Cross, D.J. Joseph, J. Cant, S.G. Cooper and J.W. Denham

Tangential breast irradiation: simple improvements

Int. J. Radiation Oncology Biol. Phys. 23: 433-442 (1992)

[Cunningham 1984] J.R. Cunningham

Quality assurance in dosimetry and treatment planning

Int. J. Radiation Oncology Biol. Phys. 10, Suppl.1: 105-109 (1984)

[Davis 1997]

J.B. Davis, A. Pfafflin and A. Fogliata-Cozzi

Accuracy of two - and three-dimensional photon dose calculation for

tangential breast irradiation of the breast

156

Radiother. Oncol. 42: 245-248 (1997)

[Dawes 1988]

P.J.D.K. Dawes, E.G. Aird and I.P. Crawshaw

Direct measurement of dose at depth in breast cancer using lithium

fluoride

Clinical Radiology 39: 301-304 (1988)

[de Boer 1993]

S.F. de Boer, A.S. Beddar and J.A. Rawlinson

Optical filtering and spectral measurements of radiation-induced light in

plastic scintillation dosimetry

Phys. Med. Biol. 38: 945-958 (1993)

[Ding 1995a]

W. Ding, J.E. Verstraete and J.M. Van Dam

Performance of new Scanditronix diodes for in vivo dosimetry

Acta-Oncol. 34: 268-270 (1995)

[Ding 1995b]

W. Ding, W. Patterson, L. Tremethick, D. Joseph

Calibration of entrance dose measurement for an in vivo dosimetry

programme

Australas Radiol. 39: 369-374 (1995)

[Dische 1993]

S. Dische, M.I. Saunders, C. Williams, A. Hopkins and E. Aird

Precision in reporting the dose given in a course of radiotherapy

Radiother. Oncol. 29: 287-293 (1993)

[Dixon 1982]

R.L. Dixon and K.E. Ekstrand

Silicon diode dosimetry

Int. J. Appl. Radiat. Isot. 33: 1171-1176 (1982)

[Duggan 1997]

L. Duggan, T. Kron, S. Howlett, A. Skov and P. O’Brien

An independent check of treatment plan, prescription and dose

calculation as a QA procedure

Radiother. Oncol. 42: 297-301 (1997)

157

[Dutreix 1992]

A. Dutreix, G. Leunens, J. Van Dam and E. van der Schueren

Quality assurance in radiotherapy at the patient level

AMPI Medical Physics Bulletin 17: 22-25 (1992)

[Dutreix 1997]

A. Dutreix, B.E. Bjärngard, A. Bridier, B. Mijnheer, J.E. Shaw and H.

Svensson

Monitor unit calculation for high energy photon beams

ESTRO Booklet nr 3, Garant (1997)

[Edwards 1997]

C.R. Edwards, M.H. Grieveson, P.J. Mountford and P. Rolfe

A survey of current in vivo radiotherapy dosimetry practice

Br. J. of Radiol. 70: 299-302 (1997)

[Elliott 1999]

P.A. Elliott

In vivo diode dosimetry in pelvic radiotherapy treatments using 16 MV

photons

M.Sc. thesis, University of Edinburgh (1999)

[Elliott 2001]

P.A. Elliott, L. Brown and D.I. Thwaites

In vivo dosimetry of standard and conformal pelvic radiotherapy

treatments

submitted to Radioth. Oncol.

[Essers 1993]

M. Essers, J.H. Lanson and B.J. Mijnheer

In vivo dosimetry during conformal therapy of prostatic cancer

Radiother. Oncol. 29: 271-279 (1993)

[Essers 1994]

M. Essers, R. Keus, J.H. Lanson and B.J. Mijnheer

Dosimetric control of conformal treatment of parotid gland tumours

Radiother. Oncol. 32: 154-162 (1994)

[Essers 1995]

M. Essers, J.H. Lanson, G. Leunens, T. Schnabel and B.J. Mijnheer

The accuracy of CT -based inhomogeneity corrections and in vivo

dosimetry for the treatment of lung cancer

Radiother. Oncol. 37: 199-208 (1995)

158

[Essers 1996]

M. Essers

In vivo dosimetry in radiotherapy, development, use and evaluation of

accurate patient dose verification methods

Ph. D. Thesis, Amsterdam (1996)

[

Essers 1999

]

M. Essers, and B.J. Mijnheer

In vivo dosimetry during external photon beam radiotherapy

Int. J. Radiat. Oncol. Biol. Phys. 43: 245-259 (1999)

[Fiorino 1993]

C. Fiorino, A. del Vecchio, G.M. Cattaneo, M. Fusca, B. Longobardi, P.

Signorotto and R. Calandrino

Exit dose measurements by portal film dosimetry

Radiother. Oncol. 29: 336-340 (1993)

[Fiorino 2000]

C. Fiorino, D. Corletto, P. Mangili, S. Broggi, A. Bonini, G.M. Cattaneo,

R. Parisi, A. Rosso, P. Signorotto, E. Villa, R. Calandrino

Quality Assurance by systematic in vivo dosimetry: re sults on a large

cohort of patients

Radiother. Oncol. 56: 85-95 (2000)

[Fontenla 1996a] D.P. Fontenla, J. Curran, R. Yaparpalvi and B. Vikram

Customization of a radiation management system to support in vivo

patient dosimetry using diodes

Med. Phys. 23: 1425-1429 (1996)

[Fontenla 1996b] D.P. Fontenla, R. Yaparpalvi, C.S. Chui and E. Briot

The use of diode dosimetry in quality improvement of patient care in

radiotherapy

Med. Dosim. 21: 235-241 (1996)

[Gagnon 1979]

W.F. Gagnon and J.L. Horton

Physical factors affecting absorbed dose to the skin from cobalt -60

gamma rays and 25-MV x rays

Med. Phys. 6: 285-290 (1979)

159

[Gandola 1990a] L. Gandola, S. Siena, M. Bregni, E. Sverzellati, P. Piotti, C. Stucchi, A.

Massimo Gianni and F. Lombardi

Prospective evaluation of pulmonary function in cancer patients treated

with total body irradiation, high-dose mephalan, and autologous

hematopoietic stem cell transplantation

Int. J. Radiation Oncology Biol. Phys. 19: 743-749 (1990)

[Gandola 1990b] L. Gandola, F. Lombardi, S. Siena, M. Bregni, P. Piotti, E. Sverzellati, C.

Stucchi, G. Bonadonna, A.M. Gianni and A. Lattuada

Total body irradiation and high-dose melpahalan with bone marrow

transplantation at Istituto Nazionale Tumori Milan, Italy

Radiother. Oncol. Su ppl. 1: 105-109 (1990)

[Garavaglia 1993] G. Garavaglia, K.-A. Johansson, G. Leunens and B.J. Mijnheer

The role of in vivo dosimetry

Radiother. Oncol. 29: 281-282 (1993)

[Georg 1997]

D. Georg, E. Briot, F. Julia, D. Huyskens, U. Wolff and A. Dutreix

Dosimetric comparison of an integrated multileaf collimator vs. a

conventional collimator

Phys. Med. Biol. 42: 2285-2303 (1997)

[Georg 1999]

D. Georg, B. De Ost, M.-T. Hoornaert, P. Pilette, J. Van Dam, M. Van

Dycke and D. Huyskens

Build-up modification of commercial diodes for entrance dose

measurements in ‘higher energy’ photon beams

Radiother. Oncol. 51: 249-256 (1999)

161

[

Grusell 1984

]

E. Grusell and G. Rikner

Radiation damage induced dose rate nonlinearity in a n-type silicon

detector

Acta Radiol. Oncol. 23: 465-469 (1984)

[Grusell 1986]

E. Grusell and G. Rikner

Evaluation of temperature effects in p -type silicon detectors

Phys. Med. Biol. 31: 527-534 (1986)

[Grusell 1993]

E. Grusell and G. Rikner

Linearity with dose rate of low resistivity p-type silicon semiconductor

detectors

Phys. Med. Biol. 38: 785-792 (1993)

[Hamers 1991]

H.P. Hamers, K.-A. Johansson, J.L.M. Venselaar, P. De Brouwer, U.

Hansson and C. Moudi

Entrance and exit TL-dosimetry in the conservative treatment of breast

cancer: a pilot study for the EORTC-Radiotherapy Cooperative Group"

Radiother. Oncol. 22: 280-284 (1991)

[Hamers 1993]

H.P. Hamers, K.-A. Johansson, J.L.M. Venselaar, P. De Brouwer, U.

Hansson and C. Moudi

In vivo dosimetry with TLD in conservative treatment of breast cancer

patients treated with the EORTC protocol 22881

Acta Oncologica 32: 435-443 (1993)

[Hansson 1993]

U. Hansson, K.-A. Johansson, J.C. Horiot and J. Bernier

Mailed TL dosimetry programme for machine output check and clinical

application in the EORTC radiotherapy group

Radiother. Oncol. 29: 85-90 (1993)

[Hazuka 1993]

M.B. Hazuka, D.N. Stroud, J. Adams, G.S. Ibbott, and J.J. Kinzie

Prostatic thermoluminescent dosimeter analysis in a patient treated with

18 MV X rays through a prosthetic hip

162

Int. J. Radiation Oncology Biol. Phys. 25: 339-343 (1993)

[Heukelom 1991a] S. Heukelom, J.H. Lanson, G. van Tienhoven and B.J. Mijnheer

In vivo dosimetry during tangential breast treatment

Radiother. Oncol. 22: 269-279 (1991)

[Heukelom 1991b] S. Heukelom, J.H. Lanson and B.J. Mijnheer

Comparison of entrance and exit dose measurements using ionization

chambers and silicon diodes

Phys. Med. Biol. 36: 47-59 (1991)

[Heukelom 1992] S. Heukelom, J.H. Lanson and B.J. Mijnheer

In vivo dosimetry during pelvic treatment

Radiother. Oncol. 25: 111-120 (1992)

[Heukelom 1994] S. Heukelom, J.H. Lanson and B.J. Mijnheer

Quality assurance of the simultaneous boost technique for prostatic

cancer: dosimetric aspects

Radiother. Oncol. 30: 74-82 (1994)

[Holt 1975]

J.G. Holt, G.R. Edelstein and T. E. Clark

Energy dependence of the response of lithium fluoride TLD rods in high

energy electron fields

Phys. Med. Biol. 20: 559-570 (1975)

[Hoogenhout 1990] J. Hoogenhout, W.F.M. Brouwer, J.J.M. van Gasteren, A.

Schattenberg, T. de Witte, W.A.J. van Daal and C.A.M. Haanen

Clinical and physical aspects of total body irradiation for bone marrow

transplantation in Nijmegen

Radiother. Oncol. Suppl. 1: 118-120 (1990)

[Horiot 1993]

J.C. Horiot, J. Bernier, K.-A. Johansson, E. van der Schueren and H.

Bartelink

Minimum requirements for quality assurance in radiotherapy

Radiother. Oncol. 29: 103-104 (1993)

163

[Howlett 1996]

S.J. Howlett and T. Kron

Evaluation of p-type semiconductor diodes for in vivo dosimetry in a 6

MV X-ray beam

Australas Phys. Eng. Sci. Med. 19: 83-93 (1996)

[Howlett 1999]

S. Howlett, L. Duggan, S. Bazley, and T. Kron

Selective in vivo dosimetry in radiotherapy using p -type semiconductor

diodes: a reliable quality assurance procedure.

Med. Dosim. 24: 53-56 (1999)

[Hughes 1995]

D.B. Hughes, A.R. Smith, R. Hoppe, J.B. Owen, A. Hanlon, M. Wallace

and G.E. Hanks

Treatment planning for Hodgkin's disease: a patterns of care study

Int. J. Radiation Oncology Biol. Phys. 33: 519-524 (1995)

[Hussein 1995]

S. Hussein and E. El-Khatib

Total body irradiation with a sweeping 60cobalt beam

Int. J. Radiation Oncology Biol. Phys. 33: 493-497 (1995)

[Huyskens 1994] D. Huyskens, J. Van Dam and A. Dutreix

Midplane dose determination using in vivo measurements in

combination with portal imaging

Phys. Med. Biol. 39: 1089-1101 (1994)

[

ICRU 1976

]

ICRU

Determination of absorbed dose in patients irradiated by beams of X and

gamma rays in radiotherapy procedures

Report 24. ICRU, Bethesda, Maryland (1976)

[Indovina 1989]

P.L. Indovina, M. Benassi, G.C. Giacco, A. Primavera and A. Rosati

In vivo ESR dosimetry in total body irradiation

Strahlenther. Onkol. 165: 611-616 (1989)

164

[Jornet 1996]

N. Jornet, M. Ribas and T. Eudaldo

Calibration of semiconductor detectors for dose assessment in total

body irradiation

Radiother. Oncol. 38: 247-251 (1996)

[Jornet 2000]

N. Jornet, M. Ribas and T. Eudaldo

In vivo dosimetry: intercomparison between p-type based and n-type

based diodes for 16 to 25 MV energy range

Med. Phys. 27: 1287-1293 (2000)

[Kartha 1973]

P.K.I. Kartha, A. Chung Bin and F.R. Hendrickson

Accuracy in clinical dosimetry

Br. J. Radiol. 46: 1083 (1973)

[Kesteloot 1993a] K. Kesteloot, A. Dutreix and E. van der Schueren

A model for calculating the costs of in vivo dosimetry and portal

imaging in radiotherapy departments

Radiother. Oncol. 28: 108-117 (1993)

[Kesteloot 1993b] K. Kesteloot, A. Dutreix and E. van der Schueren

Quality assurance procedures in radiotherapy - Economic criteria to

support decision making

International Journal of Technology Assessment in Health Care 9: 274-

285 (1993)

[Kidane 1992]

G. Kidane

Evaluation of radiation diodes for in vivo dosimetry during radiotherapy

M.Sc. thesis, University of St Andrews (1992)

[Kirby 1992]

T.H. Kirby, W.F. Hanson and D.A. Johnston

165

Uncertainty analysis of absorbed dose calculations from

thermoluminescence dosimeters

Med. Phys. 19: 1427-1433 (1992)

[Klevenhagen 1978] S.C. Klevenhagen

Behaviour of p-n junction silicon radiation detectors in a temperature -

compensated direct-current circuit

Med. Phys. 5: 52-57 (1978)

[Knöös 1986]

T. Knöös, L. Ahlgren and M. Nilsson

Comparison of measured and calculated absorbed doses from tangential

irradiation of the breast

Radiother. Oncol. 7: 81-88 (1986)

[Kron 1993a]

T. Kron, A. Elliot, T. Wong, G. Showell, B. Clubb and P. Metcalfe

X-ray surface dose measurements using TLD extrapolation

Med. Phys. 20: 703-711 (1993)

[Kron 1993b]

T. Kron, M. Schneider, A. Murray and H. Mameghan

Clinical thermoluminescence dosimetry: how do expectations and results

compare?

Radiother. Oncol. 26: 151-161 (1993)

[Kubo 1985]

H. Kubo

Dosimetry of anterior chest treatment by 10 MV x-rays using a tissue

compensating filter

Radiother. Oncol. 4: 185-192 (1985)

[Kutcher 1994]

G.J. Kutcher, L. Coia, M. Gillin, W.F. Hanson, S. Leibel, R.J. Morton, J.R.

Palta, J.A. Purdy, L.E. Reinstein, G.K. Svensson, M. Weller and L.

Wingfield

166

Comprehensive QA for radiation oncology: report of AAPM Radiation

Therapy Committee Task Group 40

Med. Phys. 21: 581-618 (1994)

[Lagrange 1988] J.-L. Lagrange, S. Marcié, A. Costa, M. Héry, O. Saint-Martin and N.

Brassart

Contribution à l'optimisation des mesures in vivo par detecteurs semi -

conducteurs

in "Dosimetry in Radiotherapy". IAEA Publication (IAEA -SM-298/24,

Vienna) Vol. 2: 259-273 (1988)

[Lagrange 1994] J.-L. Lagrange and A. Noel

The role of in vivo dosimetry in radiotherapy

Bull.-Cancer-Radiotherapy 81: 456-462 (1994)

[Lambert 1983]

G.D. Lambert, W.E. Liversage, A.M. Hirst and D. Doughty

Exit dose studies in megavoltage photon therapy

Br. J. of Radiol. 56: 329-334 (1983)

[

Lanson 1999

]

J.H. Lanson, M. Essers, G.J. Meijer, A.W.H. Minken, G.J. Uiterwaal and

B.J. Mijnheer

In vivo dosimetry during conformal radiotherapy

Radiother. Oncol. 52: 51-59 (1999)

[Lathinen 1988]

T. Lahtinen, H. Puurunen, P. Simonen, A. Pekkarinen and A. Väänänen

In vivo dose measurements with new linear accelerators

in "Dosimetry in Radiotherapy", 259-277, International Atomic Energy

Agency (1988)

[

Lax 1991

]

I. Lax

167

Dosimetry of

106

Ru eye applicators with a p -type silicon detector.

Phys. Med. Biol. 36: 963-972 (1991)

[Lee 1994]

P.C. Lee, J.M. Sawicka and G.P. Glasgow

Patient dosimetry quality assurance program with a commercial diode

system

Int. J. Radiation Oncology Biol. Phys. 29: 1175-1182 (1994)

[Leunens 1990a] G. Leunens, J. Van Dam, A. Dutreix and E. van der Schueren

Quality assurance in radiotherapy by in vivo dosimetry. 1. Entrance

dose measurements, a reliable procedure

Radiother. Oncol. 17: 141-151 (1990)

[Leunens 1990b] G. Leunens, J. Van Dam, A. Dutreix and E. van der Schueren

Quality assurance in radiotherapy by in vivo dosimetry. 2. Determination

of the target absorbed dose

Radiother. Oncol. 19: 73-87 (1990)

[Leunens 1991]

G. Leunens, J. Verstraete, J. Van Dam, A. Dutreix and E. van der

Schueren

In vivo dosimetry for tangential breast irradiation: role of the equipment

in the accuracy of dose delivery

Radiother. Oncol. 22: 285-289 (1991)

[Leunens 1992a] G. Leunens, J. Verstraete, W. Van Den Bogaert, J. Van Dam, A. Dutreix

and E. van der Schueren

Human errors in data transfer during the preparation and delivery of

radiation treatment affecting the final result. "Garbage in, garbage out".

Radiother. Oncol. 23: 217-222 (1992)

[Leunens 1992b] G. Leunens, J. Verstraete, A. Dutreix and E. van der Schueren

168

Assessment of dose inhomogeneity at target level by in vivo dosimetry:

can the recommended 5% accuracy in the dose delivered to the target

volume be fulfilled in daily practice?

Radiother. Oncol. 25: 242-250 (1992)

[Leunens 1993]

G. Leunens, J. Van Dam, A. Dutreix and E. van der Schueren

Importance of in vivo dosimetry as part of a quality assurance program

in tangential breast treatments

Int. J. Radiation Oncology Biol. Phys. 28: 285-296 (1993)

[Leunens 1994]

G. Leunens, J. Verstraete, J. Van Dam, A. Dutreix and E. van der

Schueren

Experience in in-vivo dosimetry investigations in Leuven

in "Radiation dose in radiotherapy from prescription to delivery" IAEA

Report TECDOC-734, 283-289 (1994)

[Li 1995]

C. Li, L.S. Lamel and D. Tom

A patient dose verification program using diode detectors

Med. Dosim. 20: 209-214 (1995)

[Loncol 1996]

Th. Loncol, J.L. Greffe, S. Vynckier and P. Scalliet

Entrance and exit dose measurements with semiconductors and

thermoluminescent dosemeters: a comparison of methods and in vivo

results

Radiother. Oncol. 41: 179-187 (1996)

[Lööf 2001]

M. Lööf, H. Nyström and G. Rikner

Characterisation of commercial n - and p -type diodes for entrance dose in

vivo dosimetry in high energy beams

Submitted to Radiother. Oncol.

169

[

Luse 1996

]

R.W. Luse, J. Eema, T. Kwiatkowski and D. Schumacher

In vivo diode dosimetry for total marrow irradiation

Int. J. Radiat. Oncol. Biol. Phys. 36: 189-195 (1996)

[Mangili 1999]

P. Mangili, C. Fiorino, A. Rosso, G.M. Cattaneo, R. Parisi, E. Villa, R.

Calandrino

In vivo dosimetry by diode semiconductors in combination with portal

films during TBI: reporting a 5-years clinical experience

Radiother. Oncol. 52: 269-276, 1999

[Marinello 1977] G. Marinello, A. Buscaill and F. Baillet

Dégradation de l'énergie d'un faisceau fixe d'électrons de 7 MeV en vue

du traitement du mycosis fongoïde

J. Radiol. Electrol. 58: 693-699 (1977)

[Mayles 2000]

W.P. Mayles, S. Heisig and H. Mayles

Treatment verification and in vivo dosimetry

chapter 11 of Radiotherapy Physics in Practice (Williams, J.R., and

Thwaites, D.I., editors ) Oxford Univ.Press,second edition (2000)

[Meijer 2001]

G.J. Meijer, A.W. Minken, K.M. van Ingen, B. Smulders, H. Uiterwaal

and B.J. Mijnheer

Accurate in vivo dosimetry of a randomized trial of prostate cancer

irradiation.

Int. J. Radiation Oncology Biol. Phys. 49: 1409-1418 (2001)

[Meiler 1997]

R.J. Meiler and M.B. Podgorsak

Characterization of the response of commercial diode detectors used for

in vivo dosimetry

Med. Dosim. 22: 31-37 (1997)

[Mellenberg 1993] D.E. Mellenberg and S.L. Schoeppel

Total scalp treatment of mycosis fungoides: the 4x4 technique

170

Int. J. Radiation Oncology Biol. Phys. 27: 953-958 (1993)

[Mijnheer 1991]

B.J. Mijnheer, S. Heukelom, J.H. Lanson, L.J. van Battum, N.A.M. van

Bree and G. van Tienhoven

Should inhomogeneity corrections be applied during treatment planning

of tangential breast irradiation?

Radiother. Oncol. 22: 239-244 (1991)

[Mijnheer 1992]

B.J. Mijnheer

Quality assurance in radiotherapy: physical and technical aspects

Qual-Assur-Health-Care 4: 9-18 (1992)

[

Mijnheer 1994

]

B.J. Mijnheer

Possibilities and limitations of in vivo dosimetry

in “Radiation dose in radiotherapy from prescription to delivery”, IAEA

Report

TECDOC-734, edited by IAEA, 259-264 (1994).

[Millwater 1993] C.J. Millwater

The role of routine in vivo dosimetry as part of quality assurance in a

radiotherapy department: a feasibility study

M.Sc. thesis, University of Edinburgh (1993)

[Millwater 1998] C.J. Millwater, A.S. MaCleod and D.I. Thwaites

In vivo semiconductor dosimetry as part of routine quality assurance

Brit. J. Radiol. 71: 661-668 (1998)

[Mitine 1991]

C. Mitine, G. Leunens, J. Verstraete, N. Blanckaert, J. Van Dam, A.

Dutreix and E. van der Schueren

Is it necessary to repeat quality control procedures for head and neck

patients?

Radiother. Oncol. 21: 201-210 (1991)

[Muller 1990]

R. Muller-Runkel and U.P. Kalokhe

Scatter dose from tangential breast irradiation to the uninvolved breast

171

Radiology 175: 873-876 (1990)

[

Muller 1991

]

R. Muller-Runkel and S.S. Watkins

Introducing a computerized record and verify system: its impact on the

reduction of treatment errors

Med. Dosimetry 16: 19-22 (1991)

[NACP 1980]

Nordic Association of Clinical Physics (NACP)

Procedures in external radiation therapy dosimetry with electron and

photon beams with maximum energies between 1 and 50 MeV

Acta Radiol. Oncol. 19: 55-79 (1980)

[Nilsson 1985]

B. Nilsson

Electron contamination from different materials in high energy photon

beams

Phys. Med. Biol. 30: 139-151 (1985)

[Nilsson 1986]

B. Nilsson and A. Brahme

Electron contamination from photon beam collimators

Radiother. Oncol. 5: 235-244 (1986)

[Nilsson 1988]

B. Nilsson, B.I. Rudén and B. Sorcini

Characteristics of silicon diodes as patient dosemeters in external

radiation therapy

Radiother. Oncol. 11: 279-288 (1988)

[Niroomand 1991] A. Niroomand-Rad

Physical aspects of total body irradiation of bone marrow transplant

patients using 18 MV X rays

Int. J. Radiation Oncology Biol. Phys. 20: 605-611 (1991)

[

Noel 1987

]

A. Noel and P. Aletti

Les mesures in v ivo systématiques. A propos de 800 contrôles

172

J. Eur. Radiother. 8 (1987)

[

Noel 1992

]

A. Noel, P. Aletti, P. Bey, L. Malissard and I. Buccheit

L’erreur est humaine - Détection par dosimétrie in vivo.

Bull. Cancer/Radiother. 79: 307-334 (1992)

[Noel 1995]

A. Noel, P. Aletti, P. Bey and L. Malissard

Detection of errors in individual patients in radiotherapy by systematic

in vivo dosimetry

Radiother. Oncol. 34: 144-151 (1995)

[Ostwald 1995]

P.M. Ostwald, T. Kron, C.S. Hamilton and J.W. Denham

Clinical use of carbon-loaded thermoluminescent dosimeters for skin

dose determination

Int. J. Radiation Oncology Biol. Phys. 33: 943-950 (1995)

[Planskoy 1996a] B. Planskoy, A.M. Bedford, F.M. Davis, P.D. Tapper and L.T. Loverock

Physical aspects of total-body irradiation at the Middlesex Hospital

(UCL group of hospitals), London 1988-1993: I. Phantom measurements

and planning methods

Phys. Med. Biol. 41: 2307-2326 (1996)

[Planskoy 1996b] B. Planskoy, P.D. Tapper, A.M. Bedford and F.M. Davis

Physical aspects of total-body irradiation at the Middlesex Hospital

(UCL group of hospitals), London 1988-1993: II. In vivo planning and

dosimetry

Phys. Med. Biol. 41: 2327-2343 (1996)

[Podgorsak 1995] M.B. Podgorsak, J.P. Balog, C.H. Sibata and A.K. Ho

173

In vivo dosimetry using a diode detector system: regarding Lee et al.,

IJROBP 29: 1175-1182 (1994)

Int. J. Radiation Oncology Biol. Phys. 32: 556-557 (1995)

[Quast 1987]

U. Quast

Total body irradiation - review of treatment techniques in Europe

Radiother. Oncol. 9: 91-106 (1987)

[Quast 1990]

U. Quast, A. Dutreix and J.J. Broerse

Late effects of total body irradiation in correlation with physical

parameters

Radiother. Oncol. Suppl. 1: 158-162 (1990)

[Quast 1991]

U. Quast

The dose to lung in TBI

Strahlenther. Onkol. 167: 135-151 (1991)

[Redpath 1992]

A.T. Redpath, D.I. Thwaites, A. Rodger, M. Aitken and P.D. Hardman

A multidisciplinary approach to improving the quality of tangential

chest wall and breast irradiation for carcinoma of the breast

Radioth. Oncol. 23: 118-126. (1992)

[Ribas 1998]

M. Ribas, N. Jornet, T. Eudaldo, D. Carabante, M.A. Duch, M. Ginjaume,

G. Gómez de Segura and F. Sánchez-Dobaldo

Midplane dose determination during total boday irradiation using in

vivo dosimetry

Radiother. Oncol. 49: 91-98 (1998)

[Rikner 1983]

G. Rikner and E. Grusell

Effects of radiation damage on p -type silicon detectors

Phys. Med. Biol. 28: 1261-1267 (1983)

[Rikner 1984]

G. Rikner and E. Grusell

174

Radiation damage induced dose rate non-linearity in an n-type silicon

detector.

Acta Radiol. Oncol. 23: 465-469 (1984)

[Rikner 1987a]

G. Rikner and E. Grusell

General specifications for silicon semiconductors for use in radiation

dosimetry

Phys. Med. Biol. 32: 1109-1117 (1987)

[Rikner 1987b]

G. Rikner and E. Grusell

Patient dose measurements in photon fields by means of silicon

semiconductor diodes

Med. Phys. 14: 870-873 (1987)

[

Rikner 1993

]

G. Rikner

Silicon diodes as detectors in relative dosimetry of photon, electron and

proton irradiation fields

Doctoral thesis, Department of Physical Biology and Division of

Hospital Physics, Uppsala University, Sweden

[Rizzotti 1985]

A. Rizzotti, C. Compri and G.F. Garusi

Dose evaluation to patients irradiated by

Co beams, by means of direct

measurement on the incident and on the exit surfaces

Radiother. Oncol. 3: 279-283 (1985)

[Rudén 1976]

B.-I. Rudén

Evaluation of the clinical use of TLD

Acta Radiologica Therapy Physics Biology 15: 447-464 (1976)

[Sánchez 1994]

F. Sánchez-Doblado, R. Arráns, J.A. Terrón, B. Sánchez-Nieto and L.

Errazquin

Computerized semiconductor probe-based system for in vivo dosimetry

of patients undergoing high-energy radiotherapy

Med. Biol. Eng. Comput. 32: 588-592 (1994)

175

[Sánchez 1995]

F. Sánchez-Doblado, J.A. Terrón, B. Sánchez-Nieto, R. Arráns, L.

Errazquin, D. Biggs, C. Lee, L. Núnez, A. Delgado and J.L. Muniz

Verification of an on line in vivo semiconductor dosimetry system for

TBI with two TLD procedures

Radiother. Oncol. 34: 73-77 (1995)

[

Sen 1996

]

A. Sen, E.I. Parsai, S.W. McNeeley and K.M. Ayyangar

Quantitative assessment of beam perturbations caused by silicon diodes

used for in vivo dosimetry

Int. J. Radiat. Oncol. Biol. Phys. 36: 205-211 (1996)

[Shakeshaft 1999] J.T. Shakeshaft, H.M. Morgan and P.D. Simpson

In vivo dosimetry using diodes as a quality control tool: experience of 2

years and 2000 patients

Br. J. Radiol. 72: 891-895 (1999)

[Sixel 1994]

K.E. Sixel and E.B. Podgorsak

Build-up region of dose maximum of megavoltage X-ray beams

Med. Phys. 21: 411-416 (1994)

[Sjögren 1996]

R. Sjögren and M. Karlsson

Electron contamination in clinical high energy photon beams

Med. Phys. 23: 1873-1881 (1996)

[Sjögren 1998]

R. Sjögren and M. Karlsson

Influence of electron contamination on in vivo surface dosimetry for

high-energy photon beams

Med. Phys. 25: 916-921 (1998)

[Svahn 1976]

G. Svahn-Tapper

Mantle treatment: absorbed dose measurements in patients compared

with dose planning

176

Acta Radiologica Therapy Physics Biology 15: 340-356 (1976)

[Svarcer 1965]

V. Svarcer, J.F. Fowler and T.J. Deeley

Exit doses for lung fields measured by lithium fluoride

thermoluminescence

Brit. J. Radiol. 38: 785-790 (1965)

[Terrón 1994]

J.A. Terrón, F. Sánchez-Doblado, R. Arráns, B. Sánchez-Nieto and L.

Errazquin

Midline dose algorithm for in vivo dosimetry

Med. Dosim. 19: 263-267 (1994)

[Thwaites 1990]

D.I. Thwaites, G.L. Ritchie and A.C. Parker

Total body irradiation for bone marrow transplantation in Edinburgh:

techniques, dosimetry and results

Radioth. Oncol. 18 (suppl. 1): 143-145 (1990)

[Umek 1996]

B. Umek, M. Zwitter and H. Habic

Total body irradiation with translation method

Radiother. Oncol. 38: 253-255 (1996)

[Van Dam 1988]

J. Van Dam, A. Rijnders, L. Vanuytsel and H.-Z. Zhang

Practical implications of backscatter from outside the patient on the dose

distribution during total body irradiation

Radiother. Oncol. 13: 193-201 (1988)

[Van Dam 1990]

J. Van Dam, G. Leunens and A. Dutreix

Correlation between temperature and dose rate dependence of

semiconductor response; influence of accumulated dose

Radiother. Oncol. 19: 345-351 (1990)

[Van Dam 1992a] J. Van Dam, C. Vaerman, N. Blanckaert, G. Leunens, A. Dutreix and E.

van der Schueren

177

Are port films reliable for in vivo exit dose measurements?

Radiother. Oncol. 25: 67-72 (1992)

[Van Dam 1992b] J. Van Dam, G. Leunens, A. Dutreix and E. van der Schueren

Is equipment performance a factor of importance for the quality of

radiotherapy?

AMPI Medical Physics Bulletin 17: 26-30 (1992)

[

Van Dam 1994

]

J. Van Dam and G. Marinello

Methods for in vivo dosimetry in external radiotherapy

ESTRO Booklet Nr. 1, Garant (1994)

[van der Schueren]

E. van der Schueren, A. Dutreix and C. Weltens

In vivo dosimetry in clinical practice: When and What to measure? How

to correct?

ESTRO Booklet, to be published

[Van Dyk 1987]

J. Van Dyk

Dosimetry for total body irradiation

Radiother. Oncol. 9: 107-118 (1987)

[Van Dyk 1993]

J. Van Dyk, R.B. Barnett, J.E. Cygler and P.C. Shragge

Commissioning and quality assurance of treatment planning computers

Int. J. Radiation Oncology Bio l. Phys. 26: 261-273 (1993)

[Van Tienhoven 1997] G. van Tienhoven, B.J. Mijnheer, H. Bartelink and D.G. Gonzalez

Quality assurance of the EORTC Trial 22881/10882: boost versus no

boost in breast conserving therapy - an overview

Strahlenther. Onkol. 173: 201-207 (1997)

[Vanitsky 1993]

S. Vatnitsky and H. Järvinen

Application of a natural diamond detector for the measurement of

relative dose distributions in radiotherapy

Phys. Med. Biol. 38: 173-184 (1993)

178

[Van Gasteren 1991] J.J.M. Van Gasteren, S. Heuke lom, H.J. Van Kleffens, R. van der

Laarse, J.L.M. Venselaar and C.F. Westermann

The determination of phantom and collimator scatter components of the

output of mega-voltage photon beams: measurement of the collimator

scatter part with a beam-coaxial narrow cylindrical phantom

Radiother. Oncol. 20: 250-257 (1991)

[

Voordeckers 1998

]

M. Voordeckers, H. Goossens, J. Rutten and W. Van den Bogaert

The implementation of in vivo dosimetry in a small radiotherapy

department

Radiother. Oncol. 47: 45-48 (1998)

[Vrtar 1998]

M. Vrtar and N. Kovacevic

A model of in vivo dosimetry and quality assurance analysis of total

body irradiation in Zagreb

Acta Medica Croat. 52: 15-26, 1998

[Wall 1982]

B.F. Wall, C.M.H. Driscoll, J.C. Strong and E.S. Fischer

The suitability of different preparations of thermoluminescent lithium

borate for medical dosimetry

Phys. Med. Biol. 27: 1023-1034 (1982)

[Wambersie 1969] A. Wambersie, J. Dutreix and A. Dutreix

Précision dosimétrique requise en radiothérapie - consequences

concernant le choix et les performances exigées des détecteurs

Journal Belge de Radiologie 52: 94-104 (1969)

[Weaver 1995]

R.D. Weaver, B.J. Gerbi and K.E. Dusenbery

Evaluation of dose variation during total skin electron irradiation using

thermoluminescent dosime ters

179

Int. J. Radiation Oncology Biol. Phys. 33: 475-478 (1995)

[Weltens 1993]

C. Weltens, G. Leunens, A. Dutreix, J.M. Cosset, F. Eschwege and E. van

der Schueren

Accuracy in mantle field irradiations: irradiated volume and daily dose

Radiother. Oncol. 29: 18-26 (1993)

[Weltens 1994]

C. Weltens, J. Van Dam, G. Leunens, A. Dutreix and E. van der Schueren

Reliability of clinical port films for measuring dose inhomogeneities in

radiotherapy for head and neck tumours

Radiother. Oncol. 30: 167-170 (1994)

[Weltens 1998]

C. Weltens, D. Huyskens, A. Dutreix and E. van der Schueren

Assessment of dose inhomogeneities in clinical practice by film

dosimetry.

Radiother. Oncol. 49: 287-294, 1998.

[

Wierzbicki 1998

]

J.G. Wierzbicki and D.S. Waid

Large discrepancies between calculated D

max

and diode readings for

small field sizes and small SSDs of 15 MV photon beams

Med. Phys. 25: 245-246 (1998)

[

Wilkins 1997

]

D. Wilkins, X. Allen Li, J. Cygler and L. Gerig

The effect of dose rate dependence of p-type silicon detecors on linac

relative dosimetry

Med. Phys. 24: 879-881 (1997)

[WHO 1988]

WHO

Quality Assurance in Radiotherapy

Geneva (1988)

[Wolff 1998]

T. Wolff, S. Carter, K.A. Langmack, N.I. Twyman and P.P. Dendy

Characterization and use of a commercial n -type diode system

Br. J. Radiol. 7: 1168-1177 (1998)

[Zhu 1998]

T.C Zhu and J.R. Palta

Entrance in vivo dosimetry with diode detectors has been demonstrated to be a valuable

technique among the standard quality assurance methods used in a radiotherapy department.

Although its usefulness seems to be generally recognized, the additional worklo ad generated

by in vivo dosimetry is one of the factors that impedes a widespread implementation.

Especially during the initial period of establishing the technique in clinical routine, the

responsible QA person is confronted with variable tasks, such as purchasing equipment,

calibrating, defining measurement and interpretation procedures. Often, this is accompanied

by the time-consuming activities of searching through literature and contacting experienced

departments in order to gather information and define the sequence of the steps to be

undertaken.

This booklet is set up as a tool to reduce these initial efforts: it is conceived as a step -by-step

guide to implement entrance in vivo dosimetry with diodes in the clinical routine of a

radiotherapy department.

The first chapter about the preparation of the measurements contains information (including

commercial specifications) on diodes, electrometers and software. Practical guidelines for the

calibration of the diodes and the determination of correction factors are given.

The second chapter discusses the actual tasks of the responsible QA person during the initial

training period, with the emphasis on the implementation of the measurement procedure (e.g.

the training of personnel with explanation of immediate actions to be undertaken in case of

out-of-tolerance measurements)

In the third chapter, the interpretation of the measurement in relation to tolerance and action

levels is discussed and possible origins and consequences of out -of-tolerance measurements

are given.

In an additional chapter, we present an overview resulting from the evaluation of a

questionnaire on how in vivo dosimetry has been implemented in different international

centres. In the final chapter, elaborate contributions are given from five centres about

particular topics in in vivo dosimetry.

EUROPEAN SOCIETY FOR THERAPEUTIC RADIOL

Y AND ONCOL

OMINIQUE

UYSKENS

OGAERTS

ERSTRAETE

ARIKA

ÖÖF

ÅKAN

YSTRÖM

LAUDIO

IORINO

ARA

ROGGI

ÚRIA

ORNET

ONTSERRAT

IBAS

AVID

I. T

HWAITES

Sponsored by

“Europe Against Cancer”

RACTICAL

UIDELINES

MPLEMENTATION

IVO

OSIMETRY

ITH

IODES

XTERNAL

ADIOTHERAPY

ITH

HOTON

EAMS

NTRANCE

OSE

)

PHYSICS FOR CLINICAL RADIOTHERAPY

BOOKLET No. 5

ISBN 90-804532-3

RACTICAL

UIDELINES

MPLEMENT

TION

IVO

OSIMETRY

ITH

IODES

XTERNAL

ADIOTHERAP

ITH

HOTON

EAMS

NTRANCE

OSE

)

TW KAFT BOOKLET 5 11-09-2001 11:46 Pagina 1