E Chung

Investigation of three radiation detectors for accurate measurement of absorbed dose in nonstandard fields

PURPOSE To establish accurate experimental dosimetry techniques for reference dose measurements in nonstandard composite fields. METHODS A cylindrical PMMA phantom filled with water was constructed, at the center of which reference absorbed dose to water for a head and neck IMRT delivery was measured. Based on the proposed new formalism for reference dosimetry of nonstandard fields [Alfonso et al., Med. Phys. 35, 5179-5186 (2008)], a candidate plan-class specific reference (pcsr) field for a typical head and neck IMRT delivery was created on the CT images of the phantom. The absorbed dose to water in the pcsr field normalized to that in a reference 10×10cm2 field was measured using three radiation detectors: Gafchromic® EBT films, a diamond detector, and a guarded liquid-filled ionization chamber developed in-house (GLIC-03). Pcsr correction factors kQpcsr,Qfpcsr,fref were determined for five different types of air-filled ionization chambers (Exradin A12, NE2571, Exradin A1SL, Exradin A14, and PinPoint® 31006) in a fully rotated delivery and in a delivery with the same MLC settings and weights but from a single gantry angle (a collapsed delivery). RESULTS The combined standard uncertainty in measuring the correction factor kQpcsr,Qfpcsr,fref using the three dosimetry techniques was 0.3%. For all the air-filled ionization chambers and the pcsr field tested, the correction factor was not different from unity by more than ±0.8%. For the fully rotated delivery, the correction factors were in a narrow range of 0.9955-0.9986, while in the collapsed delivery, they were in a slightly broader range of 0.9922-1.0048. In the collapsed delivery, the Farmer-type chambers (Exradin A12 and NE2571) had very similar correction factors (0.9922 and 0.9931, respectively), whereas the correction factors for the smaller chambers showed more distinct chamber-type dependence. CONCLUSIONS The authors have established three experimental dosimetry techniques that allow reference measurements of nonstandard field correction factors kQpcsr,Qfpcsr,fref for air-filled ionization chambers at the 0.3% 1σ uncertainty level. These techniques can be used to determine criteria for the selection of plan-class specific reference fields and ultimately improve clinical reference dosimetry of nonstandard fields.

Dose homogeneity specification for reference dosimetry of nonstandard fields

PURPOSE To investigate the sensitivity of the plan-class specific correction factor to dose distributions in composite nonstandard field dosimetry. METHODS A cylindrical water-filled PMMA phantom was constructed at the center of which reference absorbed dose could be measured. Ten different TomoTherapy(®)-based IMRT fields were created on the CT images of the phantom. The dose distribution for each IMRT field was estimated at the position of a radiation detector or ionization chamber. The dose in each IMRT field normalized to that in a reference 10 × 10 cm(2) field was measured using a PTW micro liquid ion chamber. Based on the new dosimetry formalism, a plan-class specific correction factor k(Q(pcsr),Q) (f(pcsr),f(ref)) for each field was measured for two Farmer-type chambers, Exradin A12 and NE2571, as well as for a smaller Exradin A1SL chamber. The dependence of the measured correction factor on parameters characterizing dose distribution was analyzed. RESULTS Uncertainty on the plan-class specific correction factor measurement was in the range of 0.3%-0.5% and 0.3%-0.8% for the Farmer-type chambers and the Exradin A1SL, respectively. When the heterogeneity of the central region of the target volume was less than 5%, the correction factor did not differ from unity by more than 0.7% for the three air-filled ionization chambers. For more heterogeneous dose deliveries, the correction factor differed from unity by up to 2.4% for the Farmer-type chambers. For the Exradin A1SL, the correction factor was closer to unity due to the reduced effect of dose gradients, while it was highly variable in different IMRT fields because of a more significant impact of positioning uncertainties on the response of this chamber. CONCLUSIONS The authors have shown that a plan-class specific correction factor can be specified as a function of plan evaluation parameters especially for Farmer-type chambers. This work provides a recipe based on quantifying dose distribution to accurately select air-filled ionization chamber correction factors for nonstandard fields.

Direct absorbed dose to water determination based on water calorimetry in scanning proton beam delivery

PURPOSE The aim of this manuscript is to describe the direct measurement of absolute absorbed dose to water in a scanned proton radiotherapy beam using a water calorimeter primary standard. METHODS The McGill water calorimeter, which has been validated in photon and electron beams as well as in HDR 192Ir brachytherapy, was used to measure the absorbed dose to water in double scattering and scanning proton irradiations. The measurements were made at the Massachusetts General Hospital proton radiotherapy facility. The correction factors in water calorimetry were numerically calculated and various parameters affecting their magnitude and uncertainty were studied. The absorbed dose to water was compared to that obtained using an Exradin T1 Chamber based on the IAEA TRS-398 protocol. RESULTS The overall 1-sigma uncertainty on absorbed dose to water amounts to 0.4% and 0.6% in scattered and scanned proton water calorimetry, respectively. This compares to an overall uncertainty of 1.9% for currently accepted IAEA TRS-398 reference absorbed dose measurement protocol. The absorbed dose from water calorimetry agrees with the results from TRS-398 well to within 1-sigma uncertainty. CONCLUSIONS This work demonstrates that a primary absorbed dose standard based on water calorimetry is feasible in scattered and scanned proton beams.

Experimental analysis of general ion recombination in a liquid-filled ionization chamber in high-energy photon beams

PURPOSE To study experimentally the general ion recombination effect in a liquid-filled ionization chamber (LIC) in high-energy photon beams. METHODS The general ion recombination effect on the response of a micro liquid ion chamber (microLion) was investigated with a 6 MV photon beam in normal and SRS modes produced from a Varian® Novalis TxTM linear accelerator. Dose rates of the linear accelerator were set to 100, 400, and 1000 MU/min, which correspond to pulse repetition frequencies of 60, 240, and 600 Hz, respectively. Polarization voltages applied to the microLion were +800 and +400 V. The relative collection efficiency of the microLion response as a function of dose per pulse was experimentally measured with changing polarization voltage and pulse repetition frequencies and was compared with the theoretically calculated value. RESULTS For the 60 Hz pulse repetition frequency, the experimental relative collection efficiency was not different from the theoretical one for a pulsed beam more than 0.3% for both polarization voltages. For a pulsed radiation beam with a higher pulse repetition frequency, the experimental relative collection efficiency converged to the theoretically calculated efficiency for continuous beams. This result indicates that the response of the microLion tends toward the response to a continuous beam with increasing pulse repetition frequency of a pulsed beam because of low ion mobility in the liquid. CONCLUSIONS This work suggests an empirical method to correct for differences in general ion recombination of a LIC between different radiation fields. More work is needed to quantitatively explain the LIC general ion recombination behavior in pulsed beams generated from linear accelerators.PURPOSE To study experimentally the general ion recombination effect in a liquid-filled ionization chamber (LIC) in high-energy photon beams. METHODS The general ion recombination effect on the response of a micro liquid ion chamber (microLion) was investigated with a 6 MV photon beam in normal and SRS modes produced from a Varian(®) Novalis Tx(TM) linear accelerator. Dose rates of the linear accelerator were set to 100, 400, and 1000 MU∕min, which correspond to pulse repetition frequencies of 60, 240, and 600 Hz, respectively. Polarization voltages applied to the microLion were +800 and +400 V. The relative collection efficiency of the microLion response as a function of dose per pulse was experimentally measured with changing polarization voltage and pulse repetition frequencies and was compared with the theoretically calculated value. RESULTS For the 60 Hz pulse repetition frequency, the experimental relative collection efficiency was not different from the theoretical one for a pulsed beam more than 0.3% for both polarization voltages. For a pulsed radiation beam with a higher pulse repetition frequency, the experimental relative collection efficiency converged to the theoretically calculated efficiency for continuous beams. This result indicates that the response of the microLion tends toward the response to a continuous beam with increasing pulse repetition frequency of a pulsed beam because of low ion mobility in the liquid. CONCLUSIONS This work suggests an empirical method to correct for differences in general ion recombination of a LIC between different radiation fields. More work is needed to quantitatively explain the LIC general ion recombination behavior in pulsed beams generated from linear accelerators.

TU‐A‐BRB‐09: Ion Recombination in a Liquid‐Filled Ionization Chamber in High‐Energy Photon Beams

Purpose: To characterize the ion recombination effect of a liquid‐filled ionization chamber (LIC) in high‐energy photon beams. Methods: The ion recombination effect on the response of a PTW micro liquid ion chamber model 31018 (microLion) was investigated with a 6 MV photon beam in normal and SRS modes produced from a Varian Novalis Tx linear accelerator. Repetition rates were set to 100, 400 and 1000 MU/min, which correspond to pulse repetition frequencies of 60, 240 and 600 Hz, respectively. Polarization voltages applied to the microLion were +800 V and +400 V. The collection efficiency of the microLion as a function of dose per pulse was experimentally measured and theoretically calculated with changing polarization voltage and repetition rate. Results: The relative collection efficiency of the microLion decreased with increasing dose per pulse or pulse repetition frequency from the linear accelerator, or with decreasing polarization voltage to the microLion because of more ion recombination in the liquid. For repetition rates of 100, 400 and 1000 MU/min, the collection efficiency decreased by up to 0.62%, 0.71% and 1.9%, respectively, in a range of dose per pulse from 0.062 to 0.429 mGy/pulse with the polarization voltage of +800 V. For +400 V, the collection efficiency decrease was by up to 0.92%, 2.0% and 5.8% at 100, 400 and 1000 MU/min, respectively. For pulsed radiation with high repetition rate, the experimental relative collection efficiency for all polarization voltages tends to the theoretically calculated efficiency for continuous beams because of low ion mobility in the liquid. We proposed a correction technique for differences in recombination between different radiation fields. Conclusions: Ion recombination effects in a LIC were quantified and a practical correction method was proposed thereby enabling the use of the LIC for accurate measurements of output factors and relative doses in small and nonstandard fields.

TH-C-BRB-10: Clinical Implication of the New Dosimetry Formalism in IMRT Quality Assurance

Purpose: To apply the new dosimetry formalism [Med. Phys. 35, 5179 (2008)] to clinical IMRTquality assurance (QA). Methods: 20 different linear accelerator (Varian Clinac 21 EX)‐based clinical IMRT fields were transferred to the CTimages of a 30×30×17 cm3 Solid Water phantom to create IMRT QA fields. The phantom position was adjusted for each QA field to place the detector or chamber at the lowest dose gradient region in a virtual PTV. The reference doses in the IMRT QA and 10×10 cm2 fields were measured using a PTW micro liquid ion chamber (microLion). Based on the new dosimetry formalism, the clinical correction factor of each IMRT QA field was measured for a calibrated Exradin A12 Farmer‐type chamber in a fully‐rotated delivery and a delivery at a single gantry angle, a collapsed delivery. For each QA field, the measured dose with the correction factor was compared with a calculated dose using Analytical Anisotropic Algorithm (AAA) or Monte Carlo(MC) methods. Results: The clinical correction factor deviated from unity by up to 2.4% and 3.7% in the fully‐rotated and collapsed deliveries, respectively, depending on the dose homogeneity at the Exradin A12 collecting volume. In the fully‐rotated delivery, the measured dose with the correction factor is different from the calculated dose to within 5% and 3% for the AAA and MC, respectively. In the collapsed delivery, the discrepancy between the measured and AAA‐calculated doses was to within 8%, while it was improved to within 3.5% compared with the MC‐calculated dose. When applying the clinical correction factor, the decrease of the measured and calculated dose discrepancy is more significant for an IMRT QA field having higher dose heterogeneity. Conclusions: This work proves that the suggested dosimetry technique is effective to improve the dosimetric consistency of clinical IMRT QA.

Poster — Thur Eve — 24: Clinical application of the new dosimetry formalism for composite nonstandard beams

The IAEA-AAPM new dosimetry formalism [Med. Phys. 35, 5179 (2008)] was applied to clinical IMRT quality assurance (QA). Twenty different IMRT QA fields were created on the CT images of a 30×30×17 cm3 Solid Water™ phantom. Two Farmer-type chambers, Exradin A12 and NE2571, and a smaller Exradin A1SL ionization chamber were cross-calibrated against a reference detector, the PTW micro liquid ion chamber (microLion), in the lowest dose gradient region in each IMRT QA field delivery. Based on the new dosimetry formalism, the clinical correction factor was measured in a fully-rotated delivery and a delivery at a single gantry angle, a collapsed delivery. For the calibrated Exradin A12, the measured dose with the clinical correction factor was compared with a calculated dose using Monte Carlo (MC) methods. The clinical correction factor deviated from unity by up to 2.4% and 3.7% in the fully-rotated and collapsed deliveries, respectively, depending on the dose distribution in the chamber collecting volume. For the Exradin A1SL, the correction factor was generally closer to unity due to the reduced dose gradient on the smaller collecting volume. In the fully-rotated delivery, the measured dose with the clinical correction factor is different from the MC-calculated dose to within 4%; while the discrepancy was greater, up to 8%, in the collapsed delivery due to the much heterogeneous dose distribution in the chamber collecting volume. This work proves that the suggested dosimetry technique is effective to improve the dosimetric consistency of clinical IMRT QA.

SU‐E‐T‐420: Sensitivity of the Plan‐Class Specific Correction Factor to the Dose Distribution in Reference Dosimetry of Nonstandard Fields

Purpose: To analyze the sensitivity of a plan‐class specific correction factor to dose distribution in nonstandard field dosimetry. Methods: Ten different TomoTherapy‐based IMRT fields were created on a water‐filled PMMA cylindrical phantom at the center of which the absorbed dose to water was measured. Dose distribution of each IMRT field was evaluated using three parameters: modulation factor (MF), modified conformal index (COIN) and homogeneity index (HI). The relative dose in each IMRT field was measured using a PTW micro liquid ion chamber. Based on the new dosimetry formalism, the plan‐class specific correction factor was measured for two 0.6 cm3 Farmer‐type chambers (Exradin A12 and NE2571) and a 0.06 cm3 Exradin A1SL chamber. We investigated the dependence of the measured correction factor on the three plan evaluation parameters. Results: Uncertainty on the measurement of the plan‐class specific correction factor was 0.2–0.4 % and 0.2–0.7 % for the Farmer‐type chambers and Exradin A1SL, respectively. When the dose homogeneity was better than 5 % along with values of MF and modified COIN smaller than 2.0 and 0.025, respectively, the correction factor of the IMRT field was not different from unity by more than 0.7 %. For more heterogeneous IMRT fields, the correction factor deviated from unity by up to 2.4 % for the Farmer‐type chambers because of increasing residual gradient effects. For the Exradin A1SL, while it was closer to unity, the correction factor was highly variable in different IMRT fields due to a more significant impact of positioning uncertainties on the response of the smaller chamber. Conclusions: We have shown that the plan‐class specific correction factor can be characterized as a function of plan evaluation parameters, especially for Farmer‐type chambers. This work provides a recipe based on quantifying dose distribution to accurately select air‐filled ionization chamber correction factors for nonstandard fields.

Poster — Thur Eve — 64: A Water Calorimetry‐Based Dosimetry Standard for Direct Measurement of Absolute Absorbed Dose in Scanning Proton Beam Delivery

The aim of this work is to develop a novel water calorimetry‐based primary standard in scanned proton beams. A relatively homogeneous dose distribution was painted with 15 layers of proton energies ranging between 128–150 MeV. A 4°C Domen‐type stagnant water calorimeter with a parallel‐plate vessel was used to directly measure the absolute absorbed dose to water D w on the central axis of the beam, at a point of minimal dose non‐uniformity (peak‐to‐trough variation < 0.25%). With the water calorimeter, the dose contributions of the individual layers to the point of measurement as well the total dose delivered were analyzed. In order to validate the calorimeter, absolute dosimetry in scattered beams (250 MeV) was performed and all calorimetric results were compared with D w determined using an Exradin T1 Mini Shonka chamber following TRS‐398 protocol. The overall 1‐sigma uncertainty on the absorbed dose determination was 0.38% (scattered) and 0.64% (scanned), compared to an estimated 1.9% uncertainty obtained with TRS‐398. The agreement between the calorimetric results and TRS‐398‐based ion chamber results was better than 0.14% (scattered) and 0.32% (scanned). The feasibility of water calorimetry in proton therapy in general and scanned beam delivery in particular has been shown both experimentally and numerically. Relative to current protondosimetry protocols (such as TRS‐398), not only would a water calorimetry‐based primary standard measure the absolute D w directly, but it would also improve the uncertainty on the overall dose measurement by a factor of 3.

TH‐C‐BRB‐06: Advanced Dosimetry Techniques for Accurate Dose Measurement of Small and Nonstandard Fields

Purpose: To establish reference dosimetry techniques for accurate dose measurement of small and nonstandard fields and application to nonstandard field deliveries. Methods and Materials: A cylindrical PMMA phantom filled with water was constructed in the center of which reference absorbed dose to water was measured. Two candidate plan‐class specific reference (pcsr) fields for (i) linac‐based and (ii) TomoTherapy®‐based typical head and neck IMRT deliveries were created on the CTimages of the phantom. The absorbed dose in each pcsr field normalized to that in a 10×10 cm2 was measured using four reference detectors: Gafchromic® EBT films a diamonddetector and an in‐house developed guarded liquid ionization chamber (GLIC‐03) for the linac‐based IMRT delivery and a PTW microLion chamber for the TomoTherapy®‐based IMRT delivery. Based on the new dosimetry formalism pcsr correction factors k Q pcsr,Q f pcsr,f ref were determined for five air‐filled ionization chambers: Exradin A12 NE2571 Exradin A1SL Exradin A14 and PinPoint® 31006. The correction factor measurements were carried out in fully‐rotated and collapsed deliveries for the linac‐based IMRT delivery and only in a fully‐rotated delivery for the TomoTherapy®‐based IMRT delivery. Results: For the linac‐based IMRT delivery the evaluated overall uncertainty in measuring k Q pcsr,Q f pcsr,f ref was 0.3 %. The k Q pcsr,Q f pcsr,f ref is chamber independent within uncertainty (0.9955–0.9986) and systematically smaller than unity in the fully‐rotated delivery. In the collapsed delivery the correction factor was more dependent on the chamber type (0.9922–1.0048). For the TomoTherapy®‐based IMRT delivery k Q pcsr,Q f pcsr,f ref was above unity (1.0037–1.0076) with a 0.3 % measurement uncertainty. However the correction factor was different by 0.33 % between the Farmer‐type chambers and the smaller ionization chambers.Conclusions: The demonstrated dosimetry techniques carried out the relative dose measurements in the pcsr fields to within 0.3 % 1 uncertainty level. These dosimetry techniques will be helpful to improve dosimetric accuracy of other nonstandard field deliveries.