Characterization of a 2D ion chamber array for the verification of radiotherapy treatments

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

2005 Phys. Med. Biol. 50 3361

(http://iopscience.iop.org/0031-9155/50/14/012)

Download details:

IP Address: 152.92.171.92

The article was downloaded on 15/10/2010 at 15:58

Please note that terms and conditions apply.

View the table of contents for this issue, or go to the journal homepage for more

Home Search Collections Journals About Contact us My IOPscience

INSTITUTE OF PHYSICS PUBLISHING PHYSICS IN MEDICINE AND BIOLOGY

Phys. Med. Biol. 50 (2005) 3361–3373 doi:10.1088/0031-9155/50/14/012

Characterization of a 2D ion chamber array for theverification of radiotherapy treatments

E Spezi, A L Angelini, F Romani and A Ferri

Servizio di Fisica Sanitaria, Policlinico S.Orsola-Malpighi, via Massarenti 9, 40138 Bologna,Italy

E-mail: emiliano.spezi@physics.org

Received 18 April 2005, in final form 1 June 2005Published 6 July 2005Online at stacks.iop.org/PMB/50/3361

AbstractIn this study we investigated the characteristics of a commercial ion chamberarray and its performance in the verification of radiotherapy plans. The devicewas the 2D Array Seven29TM model (PTW, Freiburg, Germany). This is a two-dimensional detector array with 729 ionization chambers uniformly arrangedin a 27 × 27 matrix with an active area of 27 × 27 cm2. The detector short-,medium- and long-term reproducibility have been tested through an extensiveset of repeated measurements. Short-term reproducibility was well within0.2%. Medium- and long-term reproducibility were within 1%, including set-up reproducibility errors and linac output fluctuations. Dose linearity wasalso assessed. The system response to dose was verified to be linear withinthe range 2–500 MU. Output factors matched very well pinpoint chambermeasurements performed in the same experimental conditions with a maximumlocal percentage difference of 0.4%. Furthermore, the 2D Array sensitivity tomillimetric collimator positional changes and to perturbation effect of irradiatedarea was tested. The comparison with ion chamber data carried out in waterwas very satisfying. Finally, measurements of wedge-modulated fields andIMRT beam sequence matched very well ion chamber dose profiles acquired ina water tank. The extensive tests performed in this investigation show that the2D Array Seven29 is a reliable and accurate dosimeter and that it could be auseful tool for the quality assurance and the verification of radiotherapy plans.

(Some figures in this article are in colour only in the electronic version)

1. Introduction

The verification of radiotherapy treatment plans is a very important step in complexradiotherapy techniques. The aim of radiotherapy treatment planning is to provide the bestdose conformation to the target volume, while sparing critical organs and healthy tissues.Recent advances in radiation therapy have contributed to accomplishing this goal by raising

0031-9155/05/143361+13$30.00 © 2005 IOP Publishing Ltd Printed in the UK 3361

3362 E Spezi et al

R1, C1

TARGET

GANTRY

R27, C1

Figure 1. The 2D Array Seven29. The ion chambers (grey squares) cover uniformly a squaredarea of 27 × 27 cm2. The detector spacing (centre to centre) is 1 cm. The volume of each detectoris 0.5 × 0.5 × 0.5 cm3. Detector rows (R) and columns (C) are labelled sequentially as indicated.The scanning direction of the signal acquisition sequence is also shown.

the standard of deliverable treatments. In particular, the possibility of producing complexfields and dose shaping using devices such as multi leaf collimators (MLCs) has improvedconformal radiotherapy techniques and boosted the clinical implementation of intensity-modulated radiotherapy (IMRT) (Webb 2000). A number of studies (Zelefsky et al 2002,Kam et al 2003, Krueger et al 2003) shows clinical advantages in the implementation of thesenew radiation planning and delivery techniques.

The increased complexity of clinical treatments raises the need for more accurate doseverification systems and procedures (IMRTCWG 2001, Galvin et al 2004). Kutcher et al(1994) and Fraass et al (1998) provided comprehensive reports on the implementation ofstandard quality assurance (QA) programmes in radiotherapy treatment planning. However,clinical routine IMRT QA and verification require specific programmes and measurementdevices (Low et al 1998, Tsai et al 1998, Boehmer et al 2004). Jursinic and Nelms (2003) andLetourneau et al (2004) characterized and evaluated a 2D diode array for IMRT deliveryverification. Recently Amerio et al (2004) characterized a large array pixel segmentedionization chamber. However, beam characterization was carried out only for plain photonbeams and beam sizes between 5 × 5 cm2 and 12 × 12 cm2.

In this investigation we evaluate a new commercial 2D detector array for planar dosemeasurement of clinical radiation beams. A comprehensive investigation of the dosimetricproperties of the new detector has been carried out. The performances of the 2D ion chamberarray in the verification of radiotherapy plans are also presented.

2. Materials and methods

The radiation detector used in this study was the 2D Array Seven29TM (T10024) model (PTW,Freiburg, Germany). This device, developed from a 256 ion chamber array (Poppe et al2003), is a two-dimensional detector array with 729 vented ionization chambers arrangedin a 27 × 27 matrix. As shown in figure 1, the ionization chambers are equally spaced,1 cm centre to centre, and they cover an active area of 27 × 27 cm2. Each chamberhas a size of 0.5 × 0.5 × 0.5 cm3. The linear dimensions of the 2D Array Seven29 are2.2 × 30.0 × 42.0 cm3. The reference point is located at 0.5 cm from the 2D Array

Verification of radiotherapy treatments with a 2D ion chamber array 3363

95 cm

2D Array

RW3

100 cm

4.5 cm

5 cm

of measurementEffective point

Figure 2. Standard measurement set-up for the 2D ion chamber array. The 2D ion chamber arrayis sandwiched between a build-up and a backscatter layer of polystyrene (RW3). The SSD is setto 95 cm, the thickness of the build-up and backscatter material is 4.5 cm and 5 cm, respectively.The detector reference point is located at a distance of 100 cm from the radiation source.

surface. The material surrounding the vented ionization chambers is PMMA. During dataacquisition the 2D matrix of detectors is scanned left to right and bottom to top as shownin figure 1. The measurement interval, which is the time allowed for scanning the ionchamber matrix and processing the acquired data, can be varied between 400 ms and999 ms. The device allows us to measure absorbed dose to water (Gy) and absorbed doserate to water (Gy min−1) in continuous operation mode. Measurement ranges, as specified bythe manufacturer, are 200 mGy–1000 Gy and 500 mGy min−1–8 Gy min−1 for absolute doseand dose rate measurements, respectively. The 2D Array Seven29 and its multichannel arrayinterface are calibrated by the manufacturer using a certified Co60 radiation source. An on-sitefactor correcting for the quality of the beam can be calculated and supplied to the detectoracquisition software. In this work the detector array was used in absolute dose measuringmode and dose values were corrected for pressure and temperature. The correction procedurewas carried out before each measurement session. The 2D Array is controlled by the dataacquisition software MatrixScan which can export the raw 2D dataset to an ASCII file forfurther use.

The radiation source used in this investigation was an Oncor Impression IMRT + (SiemensMedical Solutions, Concord, CA) linear accelerator equipped with an 82 leaf double focusOPTIFOCUSTM MLC operating at 6 MV. For this energy the pulse rate of our linac can be setto 50 and 300 MU min−1. In this work we used a fixed pulse rate of 300 MU min−1 which isthe pulse rate used in clinical practice.

The 2D Array measurement set-up used throughout this work is shown in figure 2. Forthe build-up and the backscatter material, we used a set of RW3 polystyrene plates (PTW,Freiburg, Germany). The RW3 slab phantom consists of 33 plates machined to 30 × 30 cm2

of various thickness. The mass density of RW3 is 1.045 g cm−3 and the electron density hasa factor relative to water of 1.012. The total thickness of the build-up was 4.5 cm whereasthe backscatter material was 5.0 cm thick. The source to surface distance (SSD) was set at95.0 cm. The 2D Array Seven29 reference point of measurement was then located at 100 cmfrom the radiation source. This ensured that the projection of the MLC leaves matched the rowof ion chambers in the 2D Array. We will refer to this measurement set-up as standard set-up.

3364 E Spezi et al

In this work the matrix datasets, as acquired by the 2D Array software, were saved to file,imported and analysed within the Matlab environment (The MathWorks Inc., Natick, MA),using some especially developed software.

2.1. Reproducibility

The performance of the 2D Array Seven29 was checked over a period of usage of a few months.A reproducibility test was designed in order to verify the long-term (months), medium-term(days to weeks) and short-term (hours or less) reproducibility of the output of the detector.The test was carried out with the 2D Array in the standard set-up (cf figure 2) and by delivering100 MUs with a fixed beam size of 15 × 15 cm2. The irradiation was repeated ten times sothat average values and standard deviations could be calculated. Matrix data were then savedto file and processed off line.

2.2. Linearity

The dose linearity test was performed by irradiating the detector with a 10 × 10 cm2 6 MVfield in the measurement conditions shown in figure 2. The detector linearity was evaluatedby measuring the 2D Array Seven29 output for deliveries of 2, 3, 5, 10, 20, 50, 100, 200, 300and 500 MUs. The data presented in section 3.2 are the average of ten consecutive irradiationsfor each MU delivery.

2.3. Output factors

The performance of the 2D Array when measuring the linac radiation output as a functionof the field size was investigated. Effectively this test assessed the scatter properties of thematrix of detectors, which depend on the internal design of the device. The response of the 2DArray for small radiation beams was of particular interest because of its potential applicationsto the verification of IMRT plans. In this work output factor measurements were carried outby delivering 100 MUs for squared field sizes ranging from 2 × 2 cm2 to 27 × 27 cm2. The2D Array Seven29 was positioned in the standard set-up (cf figure 2). Dose outputs werecompared with ion chamber measurements taken in the same conditions using an RW3 platewith a special insert for a pinpoint detector (type 31014). The pinpoint detector was connectedto a Unidos universal dosimeter (PTW, Freiburg, Germany). All measurements were correctedfor temperature and pressure.

2.4. Sensitivity

As discussed in section 2, the ionization chambers of the 2D Array Seven29 are equally spaced,1 cm centre to centre. The edges of two adjacent detectors are then separated by a 0.5 cm gap.The lateral motion of electrons produced by photon scattering in the array material betweenthe detectors could lead to signal perturbation. This effect was studied for the PTW LA48linear ion chamber array by Martens et al (2001). Furthermore, the geometrical separation ofdetectors could lead to a loss of sensitivity to positional changes of the collimation system.With this test we investigated the perturbation effect of irradiated area of the 2D Array and thedetectors’ sensitivity to positional changes of the collimation system.

The 2D Array was irradiated with six rectangular and asymmetric fields (A, B, C, D, Eand F in figure 3). The beam sequence is shown in figure 3(a). The centre of the six fieldswas set to −10.0, −6.0, −2.0, +2.0, +6.0 and +10.0 cm off the x axis respectively. The fieldscentre-to-centre distance was 4 cm. The initial dimension of each field was 4 × 27 cm2.

Verification of radiotherapy treatments with a 2D ion chamber array 3365

-5 -4 -3 -2 -1 0 1X Axis (cm)

-3

-2

-1

0

1

2

3

Y A

xis

(cm

)

Y1=-3, Y2=0

-5 -4 -3 -2 -1 0 1X Axis (cm)

-3

-2

-1

0

1

2

3Y

Axi

s (c

m)

Y1=-4, Y2=0

-5 -4 -3 -2 -1 0 1X Axis (cm)

-3

-2

-1

0

1

2

3Y

Axi

s (c

m)

Y1=-2, Y2=0

-5 -4 -3 -2 -1 0 1X Axis (cm)

-3

-2

-1

0

1

2

3

Y A

xis

(cm

)

Y1=-3.5, Y2=0

A E FDC

(b)

(a)

(c)

(e)(d)

B

Figure 3. Experimental set-up for the investigation of perturbation effect of irradiated area ofthe 2D Array and of the detector sensitivity to positional changes of the collimation system. Sixrectangular and asymmetric fields with different off-axis positions are delivered (a). Initial fieldsize is 4 × 27 cm2. For each field one of the secondary collimators (Y1) is gradually moved toclose the radiation field with millimetric steps. The positions of Y1 at −4.0 cm, −3.5 cm and−3 cm are shown for field C in panels (b), (c) and (d) respectively. The final size of the radiationfield is (2 × 27) cm2 (e). Only the central part of the 2D detector is shown for clarity.

One of the collimator jaws (Y1 in figure 3) was then gradually moved to close the field withan increment step of 1 mm. The final size of each radiation field was 2 × 27 cm2. This isshown for four different collimator positions of field C in figures 3(b)–(e). At each step a newacquisition was taken by delivering 100 MUs. With this irradiation sequence, the detectoractive area was fully tested.

The 2D Array was placed in the standard set-up (cf figure 2). The sequence ofmeasurements was also taken in a water phantom with a semiflex ion chamber 0.125 cm3

(type 31010), in the same geometrical conditions. The semiflex detector was connected to anMP3 dual-channel electrometer (PTW, Freiburg, Germany). Results are shown in section 3.4.

2.5. Clinical application

The performance of the 2D Array, when measuring clinical dose maps, was also investigated.Dose profiles of 5 × 5 cm2, 10 × 10 cm2, 15 × 15 cm2 and 20 × 20 cm2 open- and wedge-modulated fields have been measured with the 2D Array in standard set-up. The results were

3366 E Spezi et al

X Axis (mm)

Y A

xis

(mm

)

−60 −40 −20 0 20 40 60

−100

−80

−60

−40

−20

0

20

40

60

80

100

10

20

30

40

50

60

70

80

90

(a)

s1 s2 s3 s4

s5 s6 s7 s8

(b)

Figure 4. Total dose map (a) and segment sequence (b) for the studied 6 MV IM beam for aprostate treatment. Dose verification measurements were performed with the 2D Array and with apinpoint chamber in a water tank. The position where the ion chamber profiles were taken is alsoshown (dashed line).

then compared with the output from a semiflex ion chamber located at 5 cm deep in a watertank with SSD = 95 cm. Furthermore, the dose delivered from a sequence of MLC modulated(IM) segments was assessed. Figure 4 shows the dose map from one of the five IM beamsoptimized by our clinical TPS (Pinnacle3, Philips Medical Systems, The Netherlands) for aprostate IMRT case. The clinical treatment field segments (figure 4(b)) were re-calculated tothe geometry of the phantom and then transferred to the linac. All the treatment parametersremained unchanged, except the gantry angle which was set to 0◦. Dose measurements werecarried out with the 2D Array in standard set-up and with a pinpoint chamber at 5 cm deep ina water tank. The detector, controlled by the Mephysto software, was connected to a Tandemelectrometer and to an MP3 beam analyser (PTW, Freiburg, Germany). All measurementswere corrected for temperature and pressure. In figure 4(a) the position of the profiles takenwith the pinpoint chamber is also depicted. The 2D Array dose map was obtained usingauto sequence delivery whereas for the ion chamber measurements each beam segment wasdelivered separately. The weighted sum of the dose output profiles was then taken.

3. Results

3.1. Reproducibility

Figure 5 shows the output of the 2D Array CAX chamber as a function of the number ofmeasurement (days). The total number of measurement points was 42, covering 4 months.Measurements were carried out in batches of ten. The experimental average of ten repeatedmeasurements with the associated standard deviation is depicted.

The reproducibility of the measurements within each batch is excellent: the maximumstandard deviation of the mean is 0.2%. This is an indicator of the very short-termreproducibility performance of the 2D Array. The total acquisition time for each batch was≈40 min. Deviations between batches are higher but still very good. The maximum dailydeviation is 0.7%. We also calculated the percentage difference between each batch averagevalue and the average of the entire set of measurements. The distribution of the data indicatesthat the medium- and long-term reproducibility of the 2D Array output is well within ±1%of the average dose value for the period of interest. We found that the reproducibility of

Verification of radiotherapy treatments with a 2D ion chamber array 3367

0 5 10 15 20 25 30 35 40 450.95

0.955

0.96

0.965

0.97

Time of use (day)

Abs

olut

e do

se (

Gy) Dec Jan

Feb

Mar

Apr

Figure 5. Reproducibility test for the 2D Array in standard set-up. Absolute CAX dose for differentirradiations is depicted. The month corresponding to the acquisition date is also indicated. Valuesare the average of ten repeated measurements. The standard deviation is also shown.

0 100 200 300 400 5000

50

100

150

200

250

300

350

400

450

500

MU

Ion

cham

ber

rela

tive

sign

al

y = 1*x – 0.0085

CAX detector linear regression

Figure 6. Linearity test for the 2D Array in standard set-up (cf figure 2). Central axis (CAX)measurements are shown as symbols. The results of the linear regression (solid line) are alsoshown. Monitor units were set to: 2, 3, 5, 10, 20, 50, 100, 200, 300 and 500.

other irradiated ion chambers was very similar to CAX chamber results and therefore are notpresented in this paper. Furthermore, deviations were well within the ±1% variations on thechamber-to-chamber response certified by the manufacturer. It must be noted that our resultsinclude linac output fluctuations and daily set-up reproducibility variations.

3.2. Linearity

The results of the linearity test are shown in figure 6. The linear regression equation is alsoshown. Although the response of the 2D Array is in absolute dose (Gy), for the purpose of thistest, we multiplied the CAX output data by the ratio (500)/d500, where 500 is the maximumMU delivered and d(500) is the CAX absolute dose at 500 MU. This effectively scaled thegraph to normalized units. The dose response of the central axis (CAX) ionization chamberis very linear with dose. The least squares fit shows that the linear relationship between MUand detector response is good down to the lowest delivered dose with a regression coefficientof 1 and an offset of −0.0085. Negative dose values are physically not meaningful, but thisresult is well within the uncertainties intrinsic to the measurements (0.4% of the lowest dosevalue d(2)) and could well be approximated to zero. Similarly to the reproducibility test,

3368 E Spezi et al

0 5 10 15 20 25 300.75

0.8

0.85

0.9

0.95

1

1.05

1.1

1.15

Field size (cm)

Out

put f

acto

r

0 5 10 15 20 25 30−0.6

−0.4

−0.2

0

0.2

0.4

0.6

0.8

1

Loca

l per

cent

age

diffe

renc

e

pinpointPTW 2D Array

lpd

Figure 7. Output factor comparison between 2D Array (◦) and pinpoint measurements (×). Thelocal percentage difference is also shown (solid line). Field sizes range between 2 × 2 cm2 and27 × 27 cm2. Data were acquired in standard set-up (cf figure 2) using the RW3 polystyrenephantom.

the analysis of the response of other chambers within the radiation field showed very similarcharacteristics of linearity and therefore are not presented in this paper.

3.3. Output factors

Output factor measurement results are shown in figure 7. The agreement between the 2DArray CAX ionization chamber values and the pinpoint dataset is very good. Discrepanciesare within ±0.5 local percentage difference (lpd) for all field sizes. Martens et al (2000)found pinpoint chamber type 31006 to over respond for large field sizes. This behaviour wasalso reported by Stasi et al (2004). This was due to the photoelectric interactions in the steelelectrode, which made the chamber over sensitive to low-energy Compton scattered photons.In this investigation we used the next generation of pinpoint chambers (type 31014) with acentral electrode in aluminium. This should significantly reduce the chamber over responseto large field sizes. Moreover, the main interest in devices such as the 2D Array is IMRTverification for which pinpoint detectors are usually regarded as an excellent choice for outputmeasurements. Since in this test only the energy spectra of the radiation field changed withfield size, these results also confirm that the response of the 2D Array is energy independent.This is important especially for small IMRT segments.

3.4. Sensitivity

Figure 8 shows the output variation of five chambers of the 2D Array when moving one ofthe secondary collimators (Y1) from −4 cm (figure 3(b)) to −2 cm (figure 3(e)). The firstacquisition (cf figure 3) is taken with chamber 1 and chamber 5 partially covered by theprojection of the collimators, whereas chambers 2–4 are inside the open portion of the field.It can be noted that the millimetric movement of the Y1 jaw is well detected by chambers1–3 which are gradually irradiated by a smaller portion of the field. Furthermore, the signalfrom chambers 4 and 5 is also influenced by the collimator position which modifies the linac’s

Verification of radiotherapy treatments with a 2D ion chamber array 3369

−4 −3.5 −3 −2.5 −20

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Collimator position (cm)

Abs

olut

e do

se (

Gy)

12345

Figure 8. 2D Array detector sensitivity test: the output of five detectors (1–5) in figure 3 is shown.The data from each detector are displayed for every position of the secondary collimator (Y). Y1jaw moves from −4 cm to −2 cm with 0.1 cm steps.

−50 0 500

0.2

0.4

0.6

0.8

1

X axis (mm)

Rel

ativ

e do

se

−50 0 500

0.2

0.4

0.6

0.8

1

X axis (mm)

Rel

ativ

e do

se

−50 0 500

0.2

0.4

0.6

0.8

1

X axis (mm)

Rel

ativ

e do

se

−50 0 500

0.2

0.4

0.6

0.8

1

X axis (mm)

Rel

ativ

e do

se

(a) (b)

(c) (d)

Figure 9. Comparison between 2D Array (•) and ionization chamber data (solid line) formeasurement profiles along the X axis of figures 3(b)–(e).

head scattered radiation component. This experiment shows that the geometrical separationof detectors does not lead to a loss of sensitivity. The detector array is indeed very sensitiveto millimetric positional changes of the collimation system. Moreover, there is no appreciableperturbation effect of irradiated area of the 2D Array. This is also confirmed by a dosimetriccomparison between the 2D Array and the semiflex ion chamber shown in figure 9 whereprofiles for the field settings in figures 3(b)–(e) are depicted. Measurements were carried outalong the X axis. Data were independently normalized to the CAX dose of the largest field,4 × 27 cm2 (figure 3(b)). The agreement between the datasets is very good. The 2D Arraymeasurements reproduce very well the ion chamber profiles for all the jaw position sequence.Discrepancies are within ±1% in the open field area. We do not present results from all

3370 E Spezi et al

−100 −50 0 50 100

20

40

60

80

100

X axis (mm)

Rel

ativ

e do

se

(a)

15o wedge

−100 −50 0 50 100

20

40

60

80

100

120

X axis (mm)

Rel

ativ

e do

se

(b)

30o wedge

−100 −50 0 50 100

50

100

150

X axis (mm)

Rel

ativ

e do

se

(c)

45o wedge

−100 −50 0 50 100

50

100

150

200

X axis (mm)

Rel

ativ

e do

se(d)

60o wedge

Figure 10. Relative dose comparison for 20 × 20 cm2 6 MV fields measured with the 2DArray (•) and ion chamber (solid line) in standard set-up. Measurements were carried out with a15◦ wedge (a), 30◦ wedge (b), 45◦ wedge (c) and 60◦ wedge (d) respectively.

the field sequences depicted in figure 3(a) because they are very similar. This test is also anintrinsic verification of the 2D Array sensitivity to the collimators position errors.

3.5. Clinical application

Figure 10 shows a comparison between wedge profiles for a 20 × 20 cm2 6 MV field measuredwith the 2D Array and ion chamber at 5 cm depth. Relative dose profiles for 15◦, 30◦, 45◦ and60◦ wedges are depicted. Data were normalized at the central axis. The agreement between2D Array profiles and ion chamber data is excellent for all wedges. The 2D Array data fit verywell the ion chamber curve with a maximum percentage difference of 1.0% lpd within the20 × 20 cm2 radiation field for the 45◦ wedge. Furthermore, the planar detector measurementsare accurate also in the penumbra region.

Figure 11 shows a comparison between wedge profiles for a 5 × 5 cm2 field. Thiscase is interesting because it shows some limitations of the 2D Array sampling capabilities.It can be noted that for this field size the dose on the thin end of the 45◦ and 60◦ wedges(figures 11(c) and (d)) is not measured. This is because the field edges actually fall betweentwo adjacent chambers. This is due to the detector design having a 1 cm spacing of theionization chambers. Nonetheless the agreement between the two datasets is very good withthe 2D Array data accurately matching the ion chamber profiles. Results for the 10 × 10 cm2

and 15 × 15 cm2 show a very similar agreement and therefore are not included.Figure 12 shows a comparison between the 2D Array dose map and pinpoint chamber

outputs for the 6 MV IM beam of figure 4 for a number of off-axis profiles. Both datasets werenormalized to their respective CAX values which were also dose of maximum. The profilecomparison shows a very good agreement between the 2D Array data and the pinpoint detectorfor all the measured off-axis profiles. The highest dose difference (3%, normalized to CAX)is found for the off-axis profile in figure 12(a). This is a low dose region undergoing strongMLC modulation (figure 12(b)). Therefore even small inaccuracies in the set-up procedure orin the MLC leaf position (within device tolerance) can explain this difference. For all the other

Verification of radiotherapy treatments with a 2D ion chamber array 3371

−100 −50 0 50 100

20

40

60

80

100

X axis (mm)

Rel

ativ

e do

se

(a)

15o wedge

−100 −50 0 50 100

20

40

60

80

100

X axis (mm)

Rel

ativ

e do

se

(b)

30o wedge

−100 −50 0 50 100

20

40

60

80

100

X axis (mm)

Rel

ativ

e do

se

(c)

45o wedge

−100 −50 0 50 100

20

40

60

80

100

X axis (mm)

Rel

ativ

e do

se(d)

60o wedge

Figure 11. Relative dose comparison for 5 × 5 cm2 6 MV fields measured with the 2D Array(•) and ion chamber (solid line) in standard set-up. Measurements were carried out with a15◦ wedge (a), 30◦ wedge (b), 45◦ wedge (c) and 60◦ wedge (d) respectively.

−100 −50 0 50 1000

20

40

60

X = −40

Y axis (mm)

Rel

ativ

e do

se (

%)

−100 −50 0 50 1000

20

40

60

80

X = −20

Y axis (mm)

Rel

ativ

e do

se (

%)

−100 −50 0 50 1000

20

40

60

80

100X = 0

Y axis (mm)

Rel

ativ

e do

se (

%)

−100 −50 0 50 1000

20

40

60

X = +30

Y axis (mm)

Rel

ativ

e do

se (

%)

(a) (b)

(c) (d)

Figure 12. Comparison of measured profiles for the 6 MV IM beam in figure 4. 2D Array (•) andpinpoint chamber profiles (solid line) for a number of off-axis profiles are shown.

profiles, discrepancies are well within 1%. Figure 12(d) shows an interesting characteristic ofthe 2D Array profile with respect to pinpoint measurements. Although the agreement betweenthe two detectors is very good, two high dose peaks are not detected by the 2D Array. This isbecause there are no ion chambers corresponding to the spatial location of the high dose peaks.Again this is due to the ion chamber spacing of 1 cm in the 2D Array. The peak-to-valley dosedifferences are 8% and 5% (normalized to CAX).

3372 E Spezi et al

4. Discussion and conclusions

In this work the dosimetric characteristics of the 2D Array Seven29 have been investigated.We also studied the detector performance when measuring open and modulated radiotherapyphoton beams. This detector was found to be light (weight 2.4 kg) and easy to use for dailyradiotherapy plan verification. The experimental positioning and set-up procedures were alsostraightforward. The average time required to arrange the system in our standard measuringset-up was ≈5 min.

The extensive measurements performed throughout this work showed that the 2D ArraySeven29 response is very reproducible in the short, medium and long term. The system linearityto dose was also verified. Output factors comparison with pinpoint chamber measurements wasalso very good. In this work output factors agreed with reference dataset for field sizes rangingfrom 2 × 2 cm2 to 27 × 27 cm2. This can be considered a very good achievement since it isnot trivial to obtain good output factor response for small radiation fields when using matricesof detectors. This depends on the 2D detector internal design and on its scattering properties.An attractive advantage of the 2D Array Seven29 is also its energy-independent response.This is an important characteristic, especially for small IMRT segments, which was confirmedby the results obtained in this study. Experiments designed to test the detector sensitivity tosmall collimator positional changes and the field perturbation effect were also carried out.The 2D Array Seven29 successfully detected millimetric positional movements of the linacsecondary collimators. Moreover, dose profiles matched very well ion chamber measurementsperformed in a water tank for open, wedge and MLC intensity-modulated beams.

As pointed out in this paper the ion chamber spacing of 1 cm may result in a limitedsampling of the radiation beam. This may be an issue if high resolution profiles are needed,e.g. in high dose gradient regions. However, this is an efficiency problem common to otherplanar detectors currently available. Electronic portal imaging devices (EPIDs) are probablythe most promising devices for IMRT verification and in vivo dosimetry (van Esch et al 2001,2004, Warkentin et al 2003). These detectors can provide very high resolution and highlyefficient planar dose maps. However, further developments are under way for the deploymentof this technique into routine clinical practice (Chin et al 2004, Louwe et al 2004, McDermottet al 2004).

With the 2D Array Seven29 the workload can be remarkably reduced when comparedto conventional verification techniques. One of the attractive points of this class of devicesis that dose distributions are acquired, shown and potentially processed on-the-fly. Theacquisition software provides a graphical environment where dose maps can be compared andevaluated. Alternatively, measurements may be exported to ASCII files and loaded in otherapplications. In this study we developed special Matlab applications to process the measureddatasets.

On the basis of the broad range of tests performed in this study, we conclude that the 2DArray is a dosimetrically accurate and sensitive tool and that it can be a useful device for QAand verification of clinical radiotherapy beams. The possibility of using the 2D Array in placeof conventional verification systems was outside the scope of this work and requires furtherinvestigation. The possibility of obtaining higher efficiency dose maps with the 2D ArraySeven29 is currently under investigation.

Acknowledgments

The authors wish to thank Dott A Peroni (Tema Sinergie S.r.l., Faenza Italy) for usefulcomments and discussion on the PTW 2D Array detector.

Verification of radiotherapy treatments with a 2D ion chamber array 3373

References

Amerio S et al 2004 Dosimetric characterization of a large area pixel-segmented ionization chamber Med. Phys.31 414–20

Boehmer D, Bohsung J, Eichwurzel I, Moys A and Budach V 2004 Clinical and physical quality assurance forintensity modulated radiotherapy of prostate cancer Radiat. Oncol. 71 319–25

Chin P W, Lewis D G and Spezi E 2004 Correction for dose-response variations in a scanning liquid ion chamberEPID as a function of linac gantry angle Phys. Med. Biol. 49 N93–103

Fraass B, Doppke K, Hunt M, Kutcher G, Starkschall G, Stern R and Van Dyke J 1998 American Association ofPhysicists in Medicine Radiation Therapy Committee Task Group 53: quality assurance for clinical radiotherapytreatment planning Med. Phys. 10 1773–829

Galvin J M, Ezzell G, Eisbrauch A, Yu C, Butler B, Xiao Y and Rosen I 2004 Implementing IMRT in clinicalpractice: a joint document of the American Society for Therapeutic Radiology and Oncology and the AmericanAssociation of Physicists in Medicine Int. J. Radiat. Oncol. Biol. Phys. 58 1616–34

IMRTCWG 2001 Intensity-modulated radiation therapy: current status and issues of interest Int. J. Radiat. Oncol.Biol. Phys. 51 880–914

Jursinic P A and Nelms B E 2003 A 2-D diode array and analysis software for verification of intensity modulatedradiation therapy delivery Med. Phys. 30 870–9

Kam M K M, Chau R M C, Suen J, Choi P H K and Teo P M L 2003 Intensity-modulated radiotherapy innasopharyngeal carcinoma: dosimetric advantage over conventional plans and feasibility of dose escalationInt. J. Radiat. Oncol. Biol. Phys. 56 145–57

Krueger E A, Fraass B A, McShan D L, Marsh R and Pierce L J 2003 Potential gains for irradiation of chest wall andregional nodes with intensity modulated radiotherapy Int. J. Radiat. Oncol. Biol. Phys. 56 1023–37

Kutcher G J et al 1994 Comprehensive QA for radiation oncology: report of AAPM Radiation Therapy CommitteeTask Group 40 Med. Phys. 21 581–618

Letourneau D, Gulam M, Yan Di, Oldham M and Wong J W 2004 Evaluation of a 2D diode array for IMRT qualityassurance Radiat. Oncol. 70 199206

Louwe R J W, McDermott L N, Sonke J-J, Tielenburg R, Wendling M, van Herk M B and Mijnheer B J 2004The long-term stability of amorphous silicon flat panel imaging devices for dosimetry purposes Med. Phys.31 2089–995

Low D A, Gerber A L, Mutic S and Purdy J 1998 Phantoms for IMRT dose distribution measurement and treatmentverification Int. J. Radiat. Oncol. Biol. Phys. 40 1231–5

Martens C, De Wagter C and De Neve W 2000 The value of the PinPoint ion chamber for characterization of smallfield segments used in intensity-modulated radiotherapy Phys. Med. Biol. 45 2519–30

Martens C, De Wagter C and De Neve W 2001 The value of the LA48 linear ion chamber array for characterizationof intensity-modulated beams Phys. Med. Biol. 46 1131–48

McDermott L N, Louwe R J, Sonke J J, van Herk M B and Mijnheer B J 2004 Dose-response and ghosting effects ofan amorphous silicon electronic portal imaging device Med. Phys. 31 285–95

Poppe B, Mehran P, Kollhoff K and Rubach A 2003 Use of a two-dimensional ionization chamber array for qualityassurance in medical linear accelerators (in German) Z. Med. Phys. 13 115–22

Stasi M, Baiotto B, Barboni G and Scielzo G 2004 The behavior of several microionization chambers in small intensitymodulated radiotherapy fields Med. Phys. 31 2792–5

Tsai J-S et al 1998 Dosimetric verification of the intensity-modulated radiation therapy of 92 patients Int. J. Radiat.Oncol. Biol. Phys. 40 1213–30

van Esch A, Depuydt T and Huyskens D P 2004 The use of an aSi-based EPID for routine absolute dosimetricpre-treatment verification of dynamic IMRT fields Radiat. Oncol. 71 223–34

van Esch A, Vanstraelen B, Verstraete J, Kutcher G and Huyskens D P 2001 Pre-treatment dosimetric verificationby means of a liquid-filled electronic portal imaging device during dynamic delivery of intensity modulatedtreatment fields Radiat. Oncol. 60 181–90

Warkentin B, Steciw S, Rathee S and Fallone B G 2003 Dosimetric IMRT verification with a flat-panel EPIDMed. Phys. 30 3143–55

Webb S 2000 Intensity Modulated Radiation Therapy (Bristol: Institute of Physics Publishing)Zelefsky J et al 2002 High-dose intensity modulated radiation therapy for prostate cancer: early toxicity and

biochemical outcome in 772 patients Int. J. Radiat. Oncol. Biol. Phys. 53 1111–6