In vivo dosimetry during conformal therapy of prostatic cancer

R ADIOTHERAPY Radiotherapy and Oncology 29 (1993) 271-279

In vivo dosimetry during conformal therapy of prostatic cancer M. Essers*, J.H. Lanson, B.J. Mijnheer The Netherlands Cancer Institute. Antoni van Leeuwenhoe!; Huis, Plesmanlaan 1.21. 1066 CX Amsterdam, The Netherlam&

(Received 16 April 1993; revision received 18 August 1993; accepted 20 August 1993)

Abstract

In vivo dose measurements were performed during the simultaneous boost technique for prostatic cancer to check the accuracy of dose calculations by a monitor unit calculation program and a three-dimensional planning system. The dose of the large field and the boost field are given simultaneously using customized 10 mm thick Roses-metal blocks in which the boost field is cut out. Following the procedure of the quality assurance protocol for this technique, the dose at the specification point has been determined by in vivo dosimetry. The measured dose was initially too high for 5 out of 16 patients, due to unexpected differences in two beams with the same nominal beam quality and a different density correction for the femoral heads; the monitor unit calculation program was therefore checked and improved. The dose at the specification point was also compared with calculations performed by a CTbased three-dimensional (3-D) planning system. The average deviation of the 3-D planning system from the measurements is 0.1% f 1.2%. Entrance, midline and exit dose values in the central axial plane, in a cranial plane and in a plane under the transmission block have also been compared with calculations performed by the 3-D treatment planning system. The measured entrance dose is, on average, 3.4% higher than the calculated dose for the AP beam and up to 5.5% for the lateral beams. Phantom measurements were performed and showed that these differences were not related to patient set-up errors. For the various midline points the difference between the measurements and the calculations is about 0.5%. No systematic deviations were found for the exit points. The deviations resulted mainly from a change in the percentage depth dose due to the tray supporting the transmission block. Our results demonstrated that for every change in the treatment procedure the validity of the dose calculation has to be rechecked. Using the CT-based 3-D treatment planning system we are able to calculate the dose very accurately for this high dosehigh precision technique. Key words: Prostatic cancer; Radiotherapy;

Quality assurance; Treatment planning; In vivo dosimetry

1. Introduction

In general, in vivo dose measurements are useful to estimate the global accuracy in dose delivery in a department, to ensure accurate dose delivery to individual patients, to estimate the dose delivered to critical organs and to verify the accuracy in dose delivery in specific treatment situations [4,13-151. In our department, in vivo dose measurements are performed during the treatment of prostatic cancer with the simultaneous boost technique, which is applied to irradiate a large regional and a bbost field simultaneously with beams of 8 MV X-

* Corresponding author. 0167-8140/93/SO6.00

SSDI

0

rays [8]. Accurate knowledge of the dose is important in this dose escalation study in order to establish the relationship between probability of local tumour control or normal tissue injury and absorbed dose [ 1,3,11]. A quality assurance protocol was developed [6] to check the accuracy of the simultaneous boost technique. The in vivo dose measurements performed by Heukelom et al. [6] showed that the dose at the specification point was, on average, 0.5% higher than the dose calculated with the monitor unit calculation program. The dose at the specification point agreed with the calculated dose to within 2.5% for all 18 patients included in that study. In the present study, the first aim of the in vivo dose measurements was to verify the accuracy of the dose delivered to each individual patient treated with the

1993 Elsevier Scientific Publishers Ireland Ltd. All rights reserved.

0167-8140(93)01269-U

M. Essers et al. / Radiother. Oncol. 29 (1993) 271-279

272

simultaneous boost technique. The second aim was to investigate the accuracy of the dose calculations performed by a CT-based 3-D planning system at the specification point, the entrance and exit dose points as well as in off-axis points, by means of in vivo dose measurements.

which is positioned in the photon beam at 67 cm from the focus. The patients were treated on a Philips SL-15 accelerator using 8 MV X-ray beams. The treatment unit is supplied with an automatic verification system to check treatment set-up parameters, such as field size, wedges, gantry and collimator rotation.

2. Materials and methods

2.2. Patient measurements

2.1. Irradiation technique

In vivo dose measurements were performed using ptype diodes (type EDP-20, Scanditronix) [ 13- 151 connected to a home-made 8-channel electrometer. The readings of the diodes are converted to the dose at the point 2 cm under the patient surface (entrance and exit dose) using the formula

The simultaneous boost technique as used in our institution has been extensively described by Lebesque et al. [8]. It is an isocentric technique and consists of an anterior-posterior (AP) open beam and two opposing lateral wedged beams with a wedge angle of 45-55”. A total dose of 70 Gy in 35 fractions is delivered at the specification point. The shape of the boost field is cut out in a 10 mm thick Roses-metal block with a transmission factor of 63.4% and conforms to the twodimensional projection of the tumour plus a margin of 13 mm (Lebesque, pers. commun.) in the beam’s eye view (BEV, Fig. 1). Fig. 2 shows in a three-dimensional view the 3 fields with the transmission blocks, together with the planning target volume (PTV), the femoral heads and CT grey-scale images in an axial and sagittal plane. The tumour is situated behind the femoral heads for the lateral fields. The transmission blocks are positioned on a 1.5 cm thick PMMA (Perspex) support,

H

P

F

Fig. 1.Beam’s eye view images of a typical AP and lateral field for the simultaneous boost technique, computed by the three-dimensional planning system. The outer lines represent the outer block edges, the inner rectangular lines represent the diaphragm field border. The cut out in the transmission block is drawn around the two-dimensional projection of the prostate with a margin of 13 mm. The diode positions for the various off-axis dose measurements are drawn: I, isocentre; D, dose specification point; V, ventral; C, cranial; and U, under the block.

Ddi& = R&&e’Nn’ ni Ci

(1)

where Rdide is the diode reading, No the calibration factor determined under reference circumstances and Ci the correction factors. The latter have to be applied for non-reference circumstances, and originate from the variation of the sensitivity of the diode with the dose per pulse and with the photon energy spectrum [4,9]. In the AP field, the exit diode was positioned under a part of the treatment couch made from carbon fibre [2], being 3.6 cm thick. The exit diode reading therefore had to be increased by a correction factor of 1.136, caused by the inverse square law and the attenuation by the carbon layers in the couch, to obtain the exit patient dose. The tray also gave rise to a correction factor, independent of tield size but dependent on source to skin distance (SSD) and varying between a value of 0.99 for SSD = 80 cm to 1.00 for SSD = 100 cm. The exit diodes were slightly displaced with regard to the central axis in order to avoid attenuation by the entrance diode. The absorbed midline dose was deduced from the measured entrance and exit doses, following the method proposed by Rizzotti et al. [16]. Apart from the absorbed dose to the specification point, three offaxis points have been measured: two cranial points (one in the boost field and one under the transmission block) and one point ventral from the dose specification point. In the beam’s eye view of Fig. 1 the position of the diodes during the routine checks and the off-axis measurements are indicated. The position of the diodes with respect to the treatment field and patient anatomy was determined from portal images. Each patient was measured during three treatment sessions. The standard deviations in the diode reading during patient measurements for the AP field were 0.6% for the entrance dose value and 1.2% for the exit dose value. For the lateral fields the values were 0.9% and 1.8%, respectively. From these values, a standard deviation of 1.0% in the AP midline dose determination and of 1.4% for the lateral midline dose determination was obtained.

IU. Essers et al. /Radiother.

Oncol. 29 (1993) 271-279

273

Fig:. 2. 3-D view of the simultaneous boost technique for prostatic cancer. The cut outs in the transmission blocks for the AP and lateral fields are shown. A central axial and sagittal plane are drawn. The sagittal plane has been reconstructed from the axial CT slices. The femoral are shown in grey and the planning target volume (PTV) in red. The yellow wire frame around the PTV represents the calculated 95”/uis surface.

2.3. Influence of the tray with shielding blocks Because of the presence of the 1.5 cm thick PMMA support and the 1 cm thick Roses-metal block the photon fluence is attenuated. The tray transmission is 93.8%, but application of this factor in calculations leads to an underestimation of the entrance dose value (e.g., 1.7% for SSD of 90 cm and 2.7% for SSD = 80 cm) due to the fact that the small angle Compton scattering process [7] is not taken into account. Primary photons scattered in the support and transmission block will reach the upper part of the patient or phantom. This effect is negligible for large tray to skin distances but becomes very important for smaller distances (Fig. 3). For lateral fields, with SSD = 80 cm, the tray to skin

distance is only 13 cm. Therefore, for the calculation of the correct midline dose value, the method of Rizzotti et al. [ 161can be applied only if a corrected entrance value is used instead of the real entrance value. The accuracy measurements of the diode system performed on phantoms showed that this procedure should be quite reliable for patient measurements. 2.4. Dose computation In this study, the diode measurements have been compared with dose calculations performed with the monitor unit calculation program, MUCP, and the calculations performed with the CT-based 3-D treatment planning system, which will be referred to as 3-D TPS.

hf. Essers et al. / Radiother.

274

Oncol. 29 (1993)

271-279

2.4.1. Monitor unit calculation program MUCP is identical to hand-calculations with look-up tables and has the advantage of quick and consistent calculations (see Appendix 1). Beam data of the previously used SL-25 accelerator are implemented in this program. The contribution of each beam to the dose at the specification point and the wedge angles of the lateral fields is determined by a treatment planning system. Using the dose value at the specification point resulting from each beam, the number of monitor units is calculated by MUCP. For the calculation the patient is assumed to be water equivalent. Afterwards, the number of monitor units is increased by a factor of 2.7%, determined by Heukelom at al. [6], to compensate for the dose reduction by the femoral heads. 2.4.2. 3-D treatment planning system The relative dose values in the patients in the central axial plane as well as in off-axis planes were calculated using the 3-D treatment planning system (Scandiplan software version V337, Scanditronix AB). The CT data of each patient were entered in the TPS. Calculations were performed firstly by assuming the patient to be water equivalent and secondly by taking into account electron densities on a pixel to pixel basis, using the equivalent path length method along ray-lines from the focus to the point of interest. The dose calculation has been done by defining a block in the field with a transmission of 63.4%. A Clarkson integration method is implemented for calculation of the depth dose of irregular fields. Beam hardening by the introduction of the wedge is taken into account. Fig. 2 shows a reconstruction of CT slices by the 3-D planning system, the 95% isodose surface encompassing the planning target volume. The number of monitor units was calculated separately using a program written in our institution (ScpMe, see Appendix l), using the relative dose values at the dose maximum of each beam, supplied by Scandiplan.

2 c'.

8’ 104

I

,

+' ,

r~

SSD-100cm

a SSD = 90cm + SSD = 80 cm 1.02 ' 5

0

10

15

20

Phantom

depth

25

30

35

(cm)

Fig. 3. Tray factor of the PMMA support and transmission block at the central beam axis as a function of SSD and depth for a IS cm x IS cm field.

The reliability of the diode measurement system, especially in off-axis planes, was checked by phantom measurements. 3. Results 3.1. Patient measurements and monitor unit calculation program The ratio of the measured dose and the dose calculated by the MUCP at the specification point was on the average 1.013 (SD = 2.3%). In this calculation a bone density dose reduction of 2.7% was applied for all

2

2

1.10

I

Specification

point

z 4

2.5. Phantom measurements To separate the accuracy of the dose calculation algorithms from the uncertainty in the dose delivery due to patient related effects, six patient treatments were simulated with a polystyrene phantom simulating the patient geometry, but assuming the patient to be water equivalent. The measurements were performed using both an ionization chamber and the diodes, thus allowing a check of the accuracy of the diode measuring system for this technique. The equipment applied for these measurements was the same as that used by Heukelom et al. [4,5]. The ionization chambers used were NE 1712, 0.6 cm3 and PTW 233642, 0.125 cm3. The dose was calculated following the procedure of the Dutch dosimetry protocol [ 121.

2 ;ij 3 0 III 0

??

. ‘_______--“_I__‘ . _____

100

A

.

8

.

ii

:

.?

.

.

____________;____I__,._.

c

:

*'. . \'

~.

~> ____________________________~~_____~~~~~~~~~

m E z :

r

090

" 1

2

L 3

” 4

5

6

1 7

Patient

I’

” 6

9

10

“‘I 11

12

13

14

15

L 16

number

Fig. 4. Ratio of the measured and calculated dose at the isocentre, using the monitor unit calculation program MUCP. Originally, the 8 MV X-ray beam data of the SL 25 accelerator with a density correction of 2.7% (W) were used. Afterwards the 8-MV X-ray beam data of the SL I5 accelerator data combined with a density correction of I .7% were implemented in MUCP (0).

hf. Essers et al. / Radiorher.

Oncol. 29 (1993) 271-279

215

patients. For 5 out of 16 patients, the dose at the specification point measured by in vivo dosimetry was more than 2.5% higher than the calculated 200 cGy (see Fig. 4). The number of monitor units was therefore adapted for these patients, to achieve the prescribed total dose of 70 Gy at the end of the treatment. Except for one patient, the measured dose was higher than the calculated dose. The calculations performed with the 3-D TPS showed that the dose reduction at the specification point, due to the electron densities of the femoral heads, was on average 1.7% (*0.5%) compared with the water equivalent situation. The beam characteristics of the SL-15 accelerator differed slightly from those of the previously used SL-25 accelerator, which mainly resulted in different percentage depth dose values. Therefore, the dose at the specification point was also calculated using a bone density dose reduction of 1.7% and the beam data of the SL-15 accelerator (see Fig. 4). The average ratio of measured and calculated dose was now 0.997 (SD = 2.3%). 3.2. Patient system

measurements

and 3-D treatment

planning

The average ratio of the measured and calculated dose values with the standard deviation are given in Table 1. For the entrance points rather large deviations exist, which is in agreement with the small angle Compton scattering process. For the points situated in the boost field, the measured dose is on the average 3.5% higher than the calculated value for the AP field and up to 5.5% higher for the lateral fields. The standard deviation is 2.0% for these points. For the entrance points under the

G-

k

110

n

Specification

c;

point

z :: ‘0 P s 3 0 z 0

100

_________________________________~~~~~~~~~~~~ . ?? ?? . . . . . . ________ ,_________________________~_________.

.

. .

‘0 ? 3

:

0.90 *

1

(

I

2

3

s 4

z

5

L 6

7

Patient

8

9

10 11 12 13 14

’ 15 16

number

Fig. 5. Ratio of the measured and calculated dose at the specification point, using the three-dimensional treatment planning system (3-D TPS).

transmission blocks, the difference between the measurements and calculations is 1% less, but the standard deviation is larger: 3.5% for the AP field and 2.6% for the lateral fields. At the specification point, the relative dose calculations performed with the 3-D TPS combined with the calculation of the number of monitor units by the ScpMe program resulted in an agreement with the measurements within O.l%, with a standard deviation of 1.2% (Fig. 5). Both the open AP field and the wedged lateral field dose calculations agree within 0.1%. For the off-axis midline points, the agreement of the 3-D TPS with the measurements is also very good.

Table 1 Ratio (average and standard deviation) of measured to calculated dose values for I6 patients Position

AP field

Lateral fields

Ratio

SD

Ratio

SD

Isocentre Midline ventral Midline cranial Midline under block

I .OOo I .002 1.002 1.005

0.014 0.014 0.013 0.025

0.999 1.003 0.997 I.012

0.018 0.017 0.025 0.020

Entrance central Entance ventral Entrance cranial Entrance under block

1.033 1.033 1.034 I.019

0.019 0.019 0.019 0.035

1.055 1.043 1.045 1.037

0.020 0.023 0.020 0.026

Exit Exit Exit Exit

1.002 I .002 0.999 I.019

0.020 0.020 0.030 0.015

0.996 0.998 1.020 1.005

0.026 0.023 0.023 0.016

central ventral cranial under block

The relative dose calculations have been performed with a three-dimensional CT-based planning system and the absolute calculations with a program (ScpMe) which uses the relative dose values at 2 cm depth.

hf. Essers et al. / Radiother.

i

Midline

off-axis

I I

Oncol. 29 (1993) 271-279

(SD = 1.2%). The good agreement of the midline dose value with the ionization chamber values shows that the method, described above, to determine the midline dose from the entrance and exit dose values, including the influence of the tray, yields consistent results. 4. Discussion

---___

__--_______A____

1

2

3

4

CMlld

’

Ventral

5

6

7

6

Patient

9

.__________

Under

plate

IO 11 12 13 14 15 16 17'16

number

Fig. 6. Ratio of the measured and calculated midline off-axis dose. Ventral (V) off-axis measurements were performed on patients 1-6, cranial (C) measurements on patients 7-12 and cranial measurements under the transmission block on patients 13-18 (See Fig. 1). Dose calculations were performed with the 3-D TPS.

The ratio of measured to calculated dose values is 1.002 for the ventral, 1.000 for the cranial and 1.007 for the midline dose under the transmission block (Fig. 6). The standard deviations are 0.9%, 1.2% and 1.4%, respectively. For the exit points, the average ratios are very close to unity, except for the lateral cranial exit point and the AP exit point under the transmission block. 3.3. Phantom measurements Six patient situations were simulated on polystyrene phantoms. Measurements were performed with diodes and ionization chambers and compared with calculations performed with the 3-D TPS. The entrance dose values were all higher than the calculated ones: l&2.0% for the AP fields and 4.0-5.0% for the lateral fields. The difference for the entrance dose values under the blocks was 1.0% less than those observed in the boost field. The transmission of the shielding block, measured in a 5 cm x 5 cm field at SSD = 100 cm, is on average 62.8% f 0.4% for the 18 patients in this study, which is relatively 1.O% less than the expected transmission. No systematic deviations were found for the midline dose values and exit dose values. These phantom measurements were also used to determine the accuracy and reproducibility of the diode measuring system compared with ionization chamber measurements. For the AP and lateral entrance points, the accuracy is within 0.1% (SD = 0.5%), for the midline dose values within 0.5% (SD = 1.0% and 1.5%, respectively) and within 0.9% for the exit dose points

4.1. Dose at specification point and MUCP A difference of 2.5% between measured and specified dose is taken as an action level to investigate the reasons for the deviation. If no set-up errors or other mistakes are made, for instance in the number of monitor units, wedge angle or gantry angle [lo], the number of monitor units is adapted in such a way that the total dose prescribed by the radiotherapist is given at the end of the treatment. This value of 2.5% is in agreement with the reproducibility of the diode system at the dose specitication point (2.5% = 2 SD). As a result of in vivo dosimetry, the number of monitor units had to be corrected for 5 out of 16 patients, while for the chosen tolerance limit of twice the standard deviation, only 1 put of 20 patients is expected to show a difference of more than 2.5%. The average ratio of measured to calculated dose of 1.013 was larger than the ratio of 1.005 for the previous series of 18 patients [6]. Because the patients are treated with an accelerator other than the one described in the study of Heukelom et al. [6], the results confirm the fact that two beams with the same nominal beam quality can have different beam characteristics. The currently employed SL- 15 accelerator was calibrated in such a way that beam differences with the previously used SL-25 accelerator were less than 0.5%. However, these were absolute differences. Locally, the relative differences could be much larger. The wedge transmission for depths larger than 10 cm was slightly larger for this accelerator. Percentage depth dose (PDD) values for open fields differed locally 1.2% for depths larger than 10 cm. Also, the change in PDD as a result of the transmission block supported by the PMMA block differed between both accelerators, due to a slightly different beam energy. A comparison between dose calculations performed by the three-dimensional planning system using CT information and assuming unit density showed that in this series of patients, on average, the dose at the isocentre given by the lateral fields is reduced by 3.4% because of the femoral heads, compared with the water situation. Because the lateral fields are responsible for about half the total dose at the isocentre, this gives a total dose reduction of 1.7%. This means that in addition, for these patients, the dose calculated using unit density has to be divided by a factor 1.017 instead of the previously derived value of 1.027 to get the dose at the specification point. This difference is probably due to the change in

hf. Essers et al. /Radiother.

211

Oncol. 29 (1993) 271-279

position of the dose specification point in the caudal direction of the body, compared with the former study. As a consequence, the data for the SL-15 accelerator are now implemented in MUCP and afterwards a dose reduction of 1.7% is applied. Recalculating the dose at the specification point resulted in an agreement within 2.5% with the measurements for all patients, and an average deviation from the measurements of 0.3%, which means that the dose in the specification point can now be predicted very accurately with MUCP. 4.2. Central plane and off-axis dose values and 3-D treatment planning system On average, the patient entrance dose is higher than calculated. Similar results were found for the phantom measurements. For the AP field, a difference of 1.7% can be ascribed to the small angle Compton scattering process, which was not taken into account in the planning system. This is due to the fact that in our planning system, only one tray transmission factor can be used per beam. The relevant factor is the dose reduction at the depth of the dose specification point. This can be improved by implementing in the TPS percentage depth dose curves measured with a tray. Another 0.7% was due to an average deviation in SSD between the actual patient set-up and the value applied during the dose calculation. Together these explain a difference of 2.4%. For the lateral fields, 2.7% can be ascribed to the scattered photons from the tray. The different SSD value also contributes 0.7% for these fields, and the implemented beam hardening effect by the wedge accounts for another 0.6% deviation. A total difference of 4.0% can therefore be explained for the lateral fields. For one patient with an exceptionally high entrance dose, the difference in SSD from the value obtained from the CT scan resulted in a 3.4% dose difference. For both AP fields and lateral fields, the deviations for the entrance dose values under the transmission blocks are less than for the other entrance dose points. This is due to the fact that the transmission of the blocks was planned to be 63.4%, as it was for the earlier series of patients, while in reality the average transmission of the Roses-metal blocks was only 62.8% for this series of patients. Therefore, for these points, two wrongly implemented data cancelled out. The dose at the specification point calculated by the 3-D TPS agreed to within 2.0% with the measured dose for all patients. The average ratio of the actual and expected dose value was 0.999 f 0.012 for this treatment. This implies that in the future relative dose calculations combined with a calculation of the number of monitor units with ScpMe can be performed for this technique without the need to perform recalculations afterwards. For the off-axis points, the agreement with the measurements was also very good. The average devia-

tion was 0.2% for the midline dose points in the boost field and 0.7% for the midline point under the transmission. block. From the EPID pictures and the reproducibility of the diode system in off-axis planes, it turned out that positioning of the diodes in the off-axis planes is as accurate as in the central axial plane. The midline and exit dose values agree very well with calculations for patients as well as phantoms. There is no systematic difference between the measurements and the calculations. Our measurements show that we are able to predict the dose very accurately with the threedimensional planning system, not only at the isocentre and other points in the central axial plane, but also in off-axis planes. 5. Conclusions Our study shows, first, that in vivo dosimetry is a useful approach to checking the dose delivery to each patient in order to give the correct dose to the target volume during high dose - high precision treatments. Second, it is a useful instrument for verifying the dose calculation method in a quality assurance protocol for a specific treatment technique. In vivo dosimetry should be performed after each change in the treatment procedure. As a result of unexpected differences in beam characteristics between two nominally identical X-ray beams, the number of monitor units had to be adjusted for more patients than before. Small differences in beam energy and accelerator head construction resulted in modifications of the wedge parameters, the percentage depth dose, the tray transmission factor and the output factor. Finally, in vivo dosimetry also proved to be very useful for checking the accuracy of a planning system. A relatively small number of phantom and patient measurements showed that the dose can be calculated very accurately with a CT-based three-dimensional planning system, in the central plane as well as in offaxis planes. In the future, this calculation method can therefore substitute the monitor unit calculation program for the simultaneous boost technique. 6. Acknowledgements We would like to thank Drs I.A.D. Bruinvis, S. Heukelom and J.V. Lebesque for the many useful discussions with respect to the planning system, in vivo dosimetry and the prostate treatment, respectively and A. Blom for his help with the Scandiplan system. This work was financially supported by the Netherlands Cancer Foundation (NKB Grant 92-40). 7. References I Brahme,A., Chavaudra,J., Landberg,T., McCullough,E., Niisslin,F., Rawlinson,A., Svensson, G. and Svensson, H. Ac-

hf. Essers et al. / Radiolher. Oncol. 29 (1993) 271-279

218 curacy requirements and quality assurance of external beam therapy with photons and electrons. Acta Oncol. Suppl. I, 1988. De Mooy, L.G., The use of carbon tibres in radiotherapy. Radiother. Oncol. 22: 140-142, 1991. Goitein, M. Calculation of the uncertainty in the dose delivered during radiation therapy. Med. Phys. 12: 608-612, 1985. Heukelom, S., Lanson, J.H. and Mijnheer, B.J. Comparison of entrance and exit dose measurements using ionization chambers and silicon diodes. Phys. Med. Biol. 36: 47-59, 1991. Heukelom, S., Lanson, J.H. and Mijnheer, B.J. In vivo dosimetry during pelvic treatment, Radiother. Oncol. 25: 111-120, 1992. Heukelom, S., Lanson, J.H. and Mijnheer, B.J. Quality assurance of the simultaneous boost technique for prostatic cancer: dosimetric aspects. Radiother. Oncol (in press). Huang, P., Chin, L.M. and Bjarngard, B.E. Scattered photons produced by beam-modifying filters. Med. Phys. 13: 57-63, 1986. Lebesque, L.V. and Keus, R. The simultaneous boost technique; the concept of relative normalized total dose. Radiother. Oncol. 22: 45-55, 1991. Leunens, G., van Dam, J., Dutreix, A. and van der Schueren, E. Quality assurance in radiotherapy by in vivo dosimetry. I. Entrance dose measurements, a reliable procedure. Radiother. Oncol. 17: 141-151, 1990.

8. Appendix

The number of monitor units required for the dose prescribed to a patient is calculated by a computer program called MUCP and by a monitor unit calculation program belonging to the 3-D treatment planning system, called ScpMe. In our institution one monitor unit is defined as the output of the accelerator which gives a dose of 1 cGy at the depth of dose maximum for a 10 cm x 10 cm field and an SSD of 100 cm.

IO Leunens, G., Verstraete, J., Van den Bogaert, W., Van Dam, J., Dutreix, A. and van der Schueren, E. 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. II Mijnheer, B.J., Battermann, J.J. and Wambersie, A. What degree of accuracy is required and can be achieved in photon and neutron therapy? Radiother. Oncol. 8: 237-252, 1987. I2 Mijnheer, B.J., Aalbers, A.H.L., Visser, A.G. and Wittkamper, F.W. Consistency and simplicity in the determination of absorbed dose to water in high-energy photon beams: a new code of practice. Radiother. Oncol. 7: 371-384, 1986. I3 Nilsson, B., Rudin, B-I. and Sorcini, B. Characteristics of silicon diodes as patient dosimeters in external radiation therapy. Radiother. Oncol. II: 279-288, 1988. I4 Rikner, G. and Grusel!, E. General specifications for silicon semiconductors for use in radiation dosimetry. Phys. Med. Biol. 32: 1109-I 117, 1987. 15 Rikner, G. and Grusell, E. Patient dose measurements in photon fields by means of silicon semiconductor detectors. Med. Phys. 14: 870-873, 1987. I6 Rizzotti, A., Compri, R. and Garusi, G.F. Dose evaluation to patients irradiated by @Co beams, by means of direct measurements on the incident and on the exit surfaces. Radiother. Oncol. 3: 279-283, 1985:

separately with MUCP. The total dose of 200 cGy is normalized at the specification point and the fraction delivered by the individual open and wedged beams is provided by the 2-D TPS. The required number of MU for each beam is calculated with the following formula:

DkY ‘Cdens

MU=

1SQL.W - PDD - FPDD- OF .B

with: 8.1.

MUCP C dens= bone density correction factor = 1.027 and

In this rather simple program, CT information cannot be included on a pixel to pixel basis and unit density is assumed in combination with an average bone density correction. Earlier measurements showed that the dose at the specification point of a patient is on average 2.7% lower than the dose in a phantom with unit density [6]. Therefore, after the calculations for the homogeneous water situation the number of monitor units is increased by a factor 1.027 for all patients. The 2-D TPS available in our institution is used to determine the beam weight and wedge angle of each beam, necessary for a satisfying dose distribution. For a treatment machine with a motorized wedge like the SL 15 and SL 25 accelerators, the number of monitor units, MU, with and without wedge has to be calculated

IsQL

=

(SAD+ 212 (SSD + 2)2

For the beam with wedge: W = wedge factor = wf. CFLD.Csso. CDSP, where wf = wedge factor at field of 10 cm x 10 cm, SSD of 100 cm and depth of 10 cm, CFLD =

correction factor for collimator field size,

219

h4. Essers et al. / Radiother. Oncol. 29 (1993) 271-279

C sso = correction factor for SSD, and C nsp = correction factor for depth of DSP, dnsp. For the beam without wedge, W = 1. PDD(DSP) = 0.634. PDD(col1. tield,DSP) + 0.366 - PDD(block opening,DSP)

DE&$= 200 cGy, the number of monitor units for each beam is calculated using the following formula:

where ~011. field = equivalent patient, and

square of collimator

field on

block field = equivalent square of block opening on patient.

F

weights as used during patient treatment are implemented in this planning system, and the relative dose distribution, normalized at the specification point, is calculated. Dose values at 2 cm depth on the central axis of each beam relative to the dose at the specification point are determined and are called D,,,. With

[(SSD + 2)(SAD + dosr)]* PDD= [(SAD + Z)(SSD + dosp)]* ’

OF = head scatter correction factor . phantom scatter correction factor . collimator exchange effect, and = Fc. Fp. CEE, and B = transmission of PMMA tray at DSP. All distances should be given in cm. 8.2. ScpMe The field sizes of the beams are determined using the beam’s eye view option of the 3-D TPS. The beam

MU=

D “Ollll. DE?; ISQL.W’.OF.B

with W’ = wf*CFtJ,‘Csso’C2, where C2 = correction factor at a depth of 2 cm. Percentage depth dose curves measured in open fields are implemented in the planning system. A depth correction is applied for wedged beams. For the calculation of the number of monitor units, where the relative dose value at 2 cm depth is used, the depth correction of the wedge factor is taken at 2 cm depth. No depth dependent tray transmission can be implemented in the planning system. Therefore, the transmission factor B at the most important depth, which is the depth of the specification point, is used.