29

CHAPTER 3

ABSOLUTE DOSE VERIFICATION USING

IONIZATION CHAMBERS OF DIFFERENT

VOLUMES IN RAPIDARC TREATMENTS

3.1 INTRODUCTION

The delivery of intensity modulated fields has been proposed by

means of a number of techniques. These include static compensators, dynamic

electrons, step and shoot or dynamic multileaf collimators (DMLC),

tomotherapy and the volumetric modulated arc delivery (RapidArc). The

latter two methods of delivery are novel. In external radiotherapy highly

accurate calculation of dose distribution is necessary for the successful

treatment of cancer. Dose discrepancies between the TPS and measurements

can arise due to inaccurate beam and component modeling, the dose

calculation algorithm, and beam delivery. The verification of the dose

delivered at a certain reference point within the patient (The International

Commission on Radiologic Units and Measurements (ICRU) report 50, 1985)

is an important step. Point dose verification should be done to correct

potential errors prior to start of treatment. This is especially true in intensity

modulated radiation therapy (IMRT) delivery, where dose delivery can be

quite complicated, with many beam angles and intricate intensity maps. To

deliver the planned dose distributions, intensity profiles are most commonly

translated into various multileaf collimated segments. High resolution

30

absolute and relative dosimetry is of great importance to evaluate the dose

distributions and also monitor unit verifications for such small segments.

Due to the complexity of RapidArc treatment planning and delivery

techniques the patient specific pre-treatment quality assurance plays a vital

role in RapidArc. Mostly IMRT plans are verified by phantom measurements

where the doses in the phantom are calculated by transferring the fluence

distribution from the patient treatment plan and the pre-treatment verification

will be done by film and ionization chambers respectively. RapidArc

generated dose distributions often have complex shapes with high gradient

regions surrounding critical patient structures. Analysis of discrepancies

between measured and calculated doses by single point measurement in high

gradient regions is a complicated task. RapidArc treatment fields consist of

small and large irregular multileaf collimator openings. So the traditional

process of Monitor unit (MU) verification is almost unfeasible, because of

large number of irregular field openings. But the independent MU checks

cannot predict the uncertainties during the actual delivery as the true delivery

depends on the condition of accelerator, which may vary with time and the

independent MU check algorithm is subjected to limitations and

approximations in their dose calculation models. Also the accuracy of these

types of measurements should be verified with other measurement techniques

before it should be widely used. Hence a point dose measurement is

commonly used. Ionization chamber based point dose measurements in a

phantom found to be the most reliable and practical technique presently used

for intensity modulated type delivery.

RapidArc has evolved towards the use of many small radiation

fields which will be defined by number of control points to increase the

resolution of the intensity map. So the pre-treatment verification requires

ionization chambers that can accurately measure the dose with millimeter

31

spatial resolution in order to minimize lack of lateral equilibrium and volume

averaging effects. One major advantage with the cylindrical ionization

chambers is the relative independence of response as a function of incident

beam angle when the beam direction is orthogonal to the chamber

longitudinal axis. This response independence makes cylindrical ionization

chambers very convenient for verifying doses delivered using coplanar beam

geometries. The characterization of the detector such as energy dependence,

the size of the collecting volume, charge leakage, and stem materials before

measurements are important as the RapidArc dosimetry conditions are

radically different from the open field chamber calibration. Different studies

were published comparing ionization chambers of various volumes for IMRT

absolute dose verifications. In this study absolute point dose were measured

for volumetric modulated arc therapy (RapidArc) treatment delivery using

five different chamber phantom combinations.

3.2 MATERIALS AND METHODS

Thirty five different RapidArc plans conforming to the clinical

standards were selected for the study. Verification plan was subsequently

created for each treatment plan with different chamber-phantom combinations

CT scanned. This includes Medtec IMRT phantom with Exradin micro

ionization chamber (0.007cm3), Medtec IMRT phantom with PTW pinpoint

chamber (0.015cm3), PTW Octavius with PTW semiflex chamber (0.125cm

3),

PTW Octavius with 2D array and indigenously made cylindrical wax

phantom with 0.6 cm3 chamber. Table 3.1 shows the chamber-phantom

combinations and electrometers/software used in the study.

32

Table 3.1 Chamber- phantom combinations and electrometer used in

this study

Chamber- Phantom

combinations Electrometer

1.

Micro ionization chamber (0.007

cm3) combined with MEDTEC

IMRT phantom

Farmer electrometer 2570/1

2.

Pinpoint chamber (0.015 cm3)

combined with MEDTEC IMRT

phantom

PTW TanSoft V 1.2

3.

Semi flex chamber (0.125 cm3)

combined with PTW Octavius CT

phantom

PTW TanSoft V 1.2

4.

2D array 729 ion chamber array

combined with PTW Octavius

LINAC phantom

PTW- Matrix Scan (S080050)

5.

Farmer chamber (0.61 cm3)

combined with Indigenously made

cylindrical wax phantom

Farmer electrometer 2570/1

CT images were taken at 1mm slice thickness by means of a

devoted CT scanner for IMRT Medtec phantom, PTW Octavius phantom and

indigenous made circular wax phantoms along with the chamber inserted in

its position. Exact couch parameters used in the treatment machine were

included while planning to reduce the errors, with panel surface having -300

HU, panel interior having -1000 HU and movable structural rails having 200

HU. Table 3.2 shows the technical specifications of the ionization chambers

used in the study.

33

Table 3.2 Technical specifications of the ionization chambers used in

the study

Ionization

Chamber

Exradin

A16

PTW

TM31014-

0193

PTW

TM31010-

1571

NE 2571

PTW 2D-

Array 729

(T10024)

Type Micro

chamber

Pin Point

Chamber

Semiflex

Chamber

Farmer

Chamber

vented

cubic ion

chamber

Array

Active

volume 0.007 cm

3 0.015 cm

3 0.125 cm

3 0.61 cm

3 0.125 cm

3

Polarizing

voltage 300V 400V 400V 300V 400V

Wall material PMMA +

Graphte

PMMA +

Graphite

PMMA +

Graphite Graphite Graphite

Calibration

Factor (ND,W)

(cGy/nC)

366.7867 228.1515 29.9138 5.37 -

Calculations were done in Eclipse treatment planning system (TPS)

version 8.6 using the Analytical Anisotropic algorithm (AAA). All the

verification plans were done without presetting any planning parameters.

Verification plans were executed on a Varian Clinac 2100 linear accelerator

equipped with multileaf collimators having 120 leaves, with centrally forty

pairs of leaves having a projection of 0.5cm at isocentre and twenty pairs

having a projection of 1cm at isocentre. The measured isocentre absolute dose

was compared with the TPS planned.

34

The ionization chambers were chosen for the study by means of

analyzing the minimum desirable chamber characteristics for exposure

measurements. The output of the linac was calibrated before measurements to

ensure the stability of output using Technical Report Series (TRS) 398

protocol. Also all the chambers were cross calibrated with a secondary

standard dosimeter at our hospital for the photon beam. The in-house

fabricated cylindrical phantom developed for this study was made of tissue

equivalent wax. Also custom water equivalent inserts were fabricated for

0.6 cm3 farmer ionization chamber that accurately matches the external

chamber dimensions. The cylindrical phantom geometry was chosen because

it is simple and very reproducible. During the measurements, all the phantoms

were positioned with its axis perpendicular to the radiation axis and its

“reference point” located at accelerator isocenter. Figure 3.1 and Figure 3.2

shows the photographs of different phantoms and ion chambers respectively

used in the study.

Figure 3.1 Photographs of the different phantoms used for the study

35

Figure 3.2 Photographs of the different chambers used for the study

3.2.1 Verification Plan Creation

The verification plans were created for each phantom-chamber

combinations with the isocenter placed exactly at the center of respective

chambers. Figure 3.3 shows the computed tomographic image of the different

phantom-chamber combinations with respective chamber inserts. The

verification plans were recomputed with unmodified fluence patterns and

transferred to the respective phantom-chamber combinations with isocenter at

the centre of the chamber volume.

36

Figure 3.3 Computed tomographic image of the different phantom-

chamber combinations with respective chamber inserts

The calculated doses from TPS at isocenter were tabulated for all

chamber- phantom combinations. For farmer ionization chamber (0.6 cm3),

the sensitive volume was delineated in the treatment planning system and the

mean dose was measured. The charge measured was converted into absolute

dose by applying suitable correction and calibration factor and the values

were tabulated. The errors that may occur during phantom delivery relate to

the effects of the MLC such as the penumbra, rounded leaf ends, intra and

interleaf leakages and the tongue and groove design, which were not taken

into account while creating verification plan.

37

3.2.2 Phantom Setup for Verification Plan

All the chambers were positioned in the phantom at the isocentre.

Also the chambers were positioned along the coronal plane passing through

the target center. A special insert were provided in the indigenously made

cylindrical wax phantom for the insertion of 0.6 cm3 chamber. Figure 3.4

shows the verification plan window in Eclipse treatment planning system for

indigenously made cylindrical wax phantom with 0.6cm3 chamber.

Figure 3.4 Verification plan window in Eclipse treatment planning

system (version 8.6) for indigenously made cylindrical wax

phantom with 0.6cm3 chamber

As different phantom chamber combinations were used, the dose

delivered to the isocentre varies. Individual correction and calibration factors

were applied for the each measurement. The measured absolute point dose

was compared with the verification plans done in the treatment planning

system. The phantoms used in this study have different dimensions; so the

source to surface distance (SSD) also varies. The SSD was kept at 85 cm for

Medtec IMRT phantom, 86 cm for PTW Octavius phantom, and 87.5 cm for

indigenously made cylindrical wax phantom with a constant source to chamber

effective point of measurement (SAD) of 100 cms for all the phantoms.

38

3.3 RESULTS

Performance of all the ionization chambers for measuring absolute

point doses was investigated and the results were compared with the fluence

measurements done using 2D array seven29 detector. Discrepancies between

calculated and measured dose distribution were found to be within the range

of experimental uncertainty for all the chamber-phantom combinations used

in this study. Absolute dose deviations with respect to TPS dose analyzed

were grouped into 3 categories. (a) micro ionization chamber compared with

semiflex and farmer ionization chamber (b) pinpoint compared with micro

ionization chamber (c) semiflex compared with farmer ionization chamber.

The percentage deviation between TPS planned and phantom measured for

different chamber-phantom combinations were shown in Figure 3.5, Figure

3.6 and Figure 3.7. Table 3.3 shows the percentage variations, mean, and

standard deviation between chambers for all the patients studied.

Figure 3.5 Percentage variation for TPS planned and measured

isocenter dose comparing 0.6cm3, microchamber and semiflex

chamber

39

Figure 3.6 Percentage variation for TPS planned and measured

isocenter dose for pinpoint and microchamber

Figure 3.7 Percentage variation for TPS planned and measured

isocentre dose for semiflex and 0.6cm3 chamber

40

Table 3.3 Results showing % variations, mean, and standard deviation

between chambers for all patients

Sl

No.

Ionization

Chamber

Maximum

negative

deviation (%)

Maximum

positive

deviation

(%)

Mean

deviation

(%)

Standard

deviation

1 Micro chamber - 4.75 1.35 - 0.92 1.7402

2 Pinpoint chamber - 4.83 1.38 - 2.02 1.5604

3 Semiflex chamber - 1.49 1.29 -0.16 0.8669

4 0.6cm3 Chamber - 1.57 2.23 0.57 0.8878

Dosimetric errors were found in all the five chambers used. The

errors may be due to steep gradient across the point of measurements.

Nevertheless, all the chambers measured the absolute dose within 5% for all

the RapidArc plans. When comparing micro ionization chamber semiflex and

farmer chambers, used in our study, micro ionization chamber shows more

deviations as compared to semiflex and farmer chambers with a maximum

variation of -4.76%, -1.49% and 2.23% respectively. When compared with

micro ionchamber, semiflex and farmer ionization chambers, percentage

variation of results varies from negative to positive, which indicates that, the

chamber with higher volume over estimates. Also farmer chamber shows

higher deviation when compared to semiflex. The deviation was found to be

less than 1% with semiflex and farmer chamber almost all the cases. Also

positive percentage deviations were observed in most of the cases with farmer

chamber. The variation in absolute dose with micro ionization chamber (0.007

cm3) was found to be less than 2% for twenty seven cases. Farmer chamber

underestimates the measured absolute dose by a maximum of 1.57%, whereas

pinpoint chamber underestimates the calculated isocentre dose by maximum

41

of 4.8%. All the results were compared with the independent fluence

measurements done with 2D seven29 ionization chamber array (Figure 3.8).

Measured fluence agrees well with that of the calculated by the TPS for 3 mm

DTA, 3% DD, for 95% of the evaluated dose points for all the cases, in the

treated volume region.

Figure3.8 Gamma analysis results using 2D array for pre treatment

quality assurance of 35 RapidArc cases

42

The degree of underestimation was found to be higher for the

smaller volume chambers. The effect of leakage on the measured charge is

relatively greater for smaller volume chambers, as the chamber sensitivity is

proportional to volume. Also the charge contribution from the small field

located at significant distances from the point of measurement may be below

the small chambers threshold and hence not detected. On the other hand, large

volume chambers used for dosimetry of conventional external fields are quite

sensitive, since higher volume chambers are long, and the electron fluence

through them may not be uniform. Under the condition of spatial fluence

uniformity, the charge collected by the large chamber may accurately

represent the absolute dose delivered by RapidArc to the point of

measurement.

3.4 DISCUSSION

The complexity of the RapidArc technique compared to IMRT

required a re-evaluation of current methodology of treatment verification.

This study investigates the effect of detector volume in the dosimetry of

RapidArc pre-treatment quality assurance. Ionization chambers have a

volumetric effect when used in radiation measurements particularly with

farmer chamber because of its large volume. The volumetric effect can be

clearly explained with respect to its position in the radiation fields. Within the

penumbra region, the chamber may behave differently inside the field or

outside the field. At these positions, the lack of charged particle equilibrium

acts in opposed ways, with more electrons coming into the chamber than

leaving it or vice versa. However, the effect is not very likely to hold for a

complete plan because adding up all the contributions from the different

segments in RapidArc tends to compensate each other as the MLC leaves are

continuously moving over the sensitive volume of the ionization chamber.

Larger volume chambers are mostly not suited to any type of intensity

43

modulated radiotherapy dosimetry because of the greater uncertainties, unless

and until dose should be averaged and mean dose should be measured instead

of point dose.

The larger chambers exhibited severe under response at the small

field's center. This is due to partial irradiation of active volume of the

chamber in transverse position. Pinpoint chamber type 31006 over - respond

for large field sizes as suggested by Martens (2000) and Stasi (2004) et al.

This was due to the photoelectric interactions in the steel electrode, which

made the chamber over-sensitive to low energy compton scattered photons.

Therefore, pinpoint chamber is primarily suitable for relative dose

measurements in small field dosimetry. Most of the chambers used in this

study have aluminium central electrode. This should significantly reduce the

chamber over-response to large field sizes and can be used for absolute

dosimetry. The response of the pinpoint chamber increases with depth and

field size as proposed by C Martens et al (2000). Also for the smaller field

openings, the volume effect of the pinpoint chamber becomes important. At

the nominal operating voltage of 400V the pinpoint type 31014 chamber

demonstrate a strong field size dependency of the polarity correction factor

and an excess of the charge collected, which can lead to underestimation of

the collection efficiency, as specified by S. Agostinelli et al (2008). An

advantage of using the 2D array is its independent energy and dose-rate

response, which is an important characteristic, especially for small IMRT

segments.

When the ionization chamber is positioned in smaller fields

(i.e. 1x1 cm2, 2x2 cm

2),the pinpoint chamber (0.015cm

3) can give accurate

readings when compared to larger volume chambers. This is because the

smaller beams will partially irradiate the bigger volume chambers. Also the

dose contribution from the remote fields may not be detected by the smaller

44

volume chamber. But the contribution can be detected by the larger volume

chambers due to high sensitivity. In RapidArc treatment delivery, the above

two cases may be possible, so there is a possibility to compensate the

deviations between the smaller and bigger volume chambers. For clinical

intensity modulated radiotherapy fields, the smaller volume chambers

sometimes show relatively greater variation when compared to larger volume

chambers. This might be because of variation in positional accuracy or stem

effect. Any small variation in positional accuracy of small volume chambers

in IMRT fields having high gradient can show large variations in the dose

measured in IMRT fields. Also Smaller volume chambers usually are more

sensitive to radiation induced leakages and charge multiplications, which are

usually negligible for large volume chambers .Since in intensity modulated

radiotherapy delivery uses small fields, there is a tendency to employ smaller

volume ionization chambers of active volume of approximately 0.1cm3 or less

for absolute dose verification. Recently it was experimentally verified that

pinpoint ion chamber with an active volume equal to 0.009 cm3 may be used

for absolute dose verification provided the area of uniform target dose

dimension greater than or equal to 1 cm and leakage corrections are taken into

account. However the use of ionization chambers for small field dosimetry

remains debatable due to the lack of electronic equilibrium across the field.

In low gradient regions the semiflex ionization chamber provides

accurate dose values for field sizes in the range 2 to 40 cm for photon beams.

At small field sizes, volume averaging, which is due to finite size of the

detector sensitive area, and modifications to electron transport, can occur.

This may be because of the non-water equivalence of the detector which can

over or under estimate the real dose values. Also Semiflex ionization with

0.125cm3 volume and type 3642 is a rigid stem chambers have a uniform

spatial resolution along all the three axes and designed for absolute dose

measurements. As the RapidArc delivery involves small subfield deliveries,

45

the ionization chamber collecting volume may be partially irradiated at a

given instant. Even if the entire volume is irradiated; the fluence distribution

along the cavity may not be uniform due to the non-uniform beam profile.

However as the entire arc is delivered, a spatial fluence uniformity would

exist in the chamber volume. So the dose measured by the chamber may not

be affected by its dimensions. However chamber sensitivity potentially drops

as the chamber collecting volume decreases. Also the chamber leakage plays

a considerable role as the sensitivity is less. Since dose is delivered in a short

time in RapidArc, the dosimetric system leakage will not affect more on the

charge collecting efficiency.

The farmer (0.6cm3) ionization chamber can give a better result

when used in large monitor unit delivery segments or in high scoring regions.

Also the errors in using a large volume chambers due to volume averaging

effect is not predominant under such conditions. So it is suggested that dose

can be averaged over the volume while using 0.6cm3 ionization chamber for

measuring RapidArc absolute dose measurements. Also farmer chamber used

with the indigenously made cylindrical wax phantom might have improved

the results and almost comparable with other chamber-phantom combinations.

The average dose calculated by the TPS for farmer chamber is dependent

upon the contoured volume which was delineated manually and can be larger

than the manufacture’s nominal chamber volume due to slice averaging

effects present in CT images. So it is suggested that least possible slice

thickness of the order of less than or equal to 1 mm is ideal for delineating the

chamber volume. It was demonstrated that different chambers respond

individually to DMLC field conditions. The discrepancy in measurements

among chambers may be mainly due to detector design and construction

material and also the different phantoms used. Also fields with moving leaves

contain a larger proportion of the scattered photons because of less sharp

penumbras due to rounded leaf edges that travel across the entire field. The

46

deviations may also be due to a high sensitivity to positioning accuracy due to

the cross-section of the beamlet. Also the field size and depth of measurement

will alter the scatter fluence reaching the detector. Correction factors can be

applied to each ionization chamber to account for fluence perturbation effects

in RapidArc delivery.

3.5 CONCLUSION

Absolute dose measurements using semiflex ionization chamber

with intermediate volume (0.125cm3) shows good agreement with the TPS

calculated among the detectors used in this study. Positioning is very

important when using smaller volume chambers as they are more sensitive to

geometrical errors within the treatment fields. Also it is suggested to average

the dose over the sensitive volume for larger volume chambers if used for

RapidArc absolute dose measurements. All the ionization chamber-phantom

combinations used in this study can be used interchangeably for routine

RapidArc patient specific quality assurance with a satisfactory accuracy for

clinical practices.