Search

A new multileaf collimator (MLC) model has been incorporated into version 7.4 of the Pinnacle radiotherapy treatment planning system (Philips Radiation Oncology Systems, Milpitas, CA). The MLC model allows for rounded MLC leaf-ends and provides separate parameters for inter-leaf transmission, intra-leaf transmission and the tongue width of the MLC leaf. In this report we detail the method followed to commission the MLC model for a Varian 120-leaf Millennium MLC (Varian Medical Systems, Palo Alto, CA, USA) for both 6 and 10 MV photons, and test the validity of the model for an IMRT field. Dose profiles in water were measured for a range of square MLC field sizes and compared to the Pinnacle computed dose profiles; in addition, the dose distribution for a series of adjacent MLC fields was measured to observe the model's behaviour along match-lines. Based on these results intra-leaf transmissions of 1.5% for 6 MV and 1.8% for 10 MV, leaf-tip radius of 12.0 cm, an inter-leaf transmission of 0.5%, and a tongue width of 0.1 cm were chosen. Using these values to compute the planar dose distribution for a 6 MV IMRT field, the new version of Pinnacle displayed improved dosimetric agreement with the dose-to-water EPID image and ion chamber measurements when compared to the old version of Pinnacle, particularly along the MLC tongue edge and across match-lines. Discrepancies of up to 5% were observed between calculated and measured doses along match-lines for both 6 MV and 10 MV photons; however, the new MLC model did predict the presence of match-lines and was a significant improvement on the previous model.

Skin-sparing properties of megavoltage photon beams are compromised by electron contamination. Higher energy beams do not necessarily produce lower surface and basal cell layer doses due to this electron contamination. For a 5 x 5 cm field size the surface doses for 6 MVp and 18 MVp X-ray beams are 10% and 7% of their respective maxima. However, at a field size of 40 x 40 cm the percentage surface dose is 42% for both 6 MVp and 18 MVp beams. The introduction of beam modifying devices such block trays can further reduce the skin-sparing advantages of high energy photon beams. Using a 10 mm perspex block tray, the surface doses for 6 MVp and 18 MVp beams with a 5 x 5 cm field size are 10% and 8%, respectively. At 40 x 40 cm, surface doses are 61% and 63% for 6 MVp and 18 MVp beams, respectively. This trend is followed at the basal cell layer depth. At a depth of 1 mm, 18 MVp beam doses are always at least 5% smaller than 6 MVp doses for the same depth at all field sizes when normalized to their respective Dmax values. Results have shown that higher energy photon beams produce a negligible reduction of the delivered dose to the basal cell layer (0.1 mm). Only a small increase in skin sparing is seen at the dermal layer (1 mm), which can be negated by the increased exit dose from an opposing field.

The Varian millennium 120 multileaf collimator has curved leaf ends. Transmission through the leaf ends generates a small asymmetric penumbral dose effect. This design can lead to hot spots between neighbouring beam segments during step and shoot IMRT dose delivery. We have observed some matchlines with film for clinical beams optimized using the pinnacle radiotherapy treatment planning system; hence we sought to verify the optimum leaf offset required to minimize the matchline effect. An in-house program was created to control the MLC leaf banks in 2 cm steps with a 2 cm gap. The gap was varied by the following offset values from 0.0 to 0.1 cm. Two types of radiographic films (Kodak EDR and XV films) and a radiochromic film (Gafchromic MD-55-2) were used to measure the optical density maps. The films were positioned in a solid water phantom perpendicular to the beam axis and irradiated at d(max) using a 6 MV photon beam. An ion chamber (IC4) was used to measure point doses for normalization in a beam umbral minima position. The relative mean peak to valley dose ratios measured with no leaf offset were 1.31, 1.30 and 1.31 for the XV, EDR2 and Gafchromic films, respectively. For a 0.07 cm gap per leaf and a performance of end leaf repeatability of 0.01 cm, the central matchline was reduced to about 1.0 for all dosimeters, with two mini-peaks measured as 1.05, 1.05 and 1.08 each side of the matchline, for XV, EDR2 and Gafchromic, respectively. The average relative dose across the umbra for this offset was XO-mat V = 1.01, EDR = 1.01 and radiochromic film = 1.02, respectively. While we expected the beam penumbral tails from segment neighbours to cause overprediction of the dose in the central valley regions due to the energy response of radiographic films, by normalizing all dosimeters to an ion chamber reading in the minimum we could not observe any major shape distortion between the radiographic film and radiochromic film results. In conclusion, relative doses measured by radiographic and radiochromic films agree well with IC4 within +/-2%.

The Varian multi-leaf collimator has a 14.5 cm leaf extension limit from each carriage. This means the target volumes in the head and neck region are sometimes too wide for standard width-modulated fields to provide adequate dose coverage. A solution is to set up asymmetric co-axial overlapping fields. This protects the MLC carriage while in return the MLC provides modulated dose blending in the field overlap region. Planar dose maps for coincident fields from the Pinnacle radiotherapy treatment planning system are compared with planar dose maps reconstructed from radiographic film and electronic portal images. The film and portal images show small leaf-jaw matchlines at each field overlap border. Linear profiles taken across each image show that the observed leaf-jaw matchlines from the accelerator images are not accounted for by the treatment planning system. Dose difference between film reconstructed electronic portal images and planning system are about 2.5 cGy in a modulated field at d(max). While the magnitude of the dose differences are small improved round end leaf modelling combined with a finer dose calculation grid may minimize the discrepancy between calculated and delivered dose.

This short note investigates the dose characteristics of a relatively new high sensitivity radiochromic film (Gafchromic HS) and compares dose and energy response to various Gafchromic film types and radiographic (EDR-2) film. The original MD-55-1 and two improved sensitivity films, MD-55-2 and HS film, were investigated for energy and dose response. Results show that the energy response of the new HS film is relatively the same as the original MD-55-1 and MD-55-2 films with a decrease in sensitivity at lower x-ray energies, with response decreasing down to approximately 0.64 (normalized to 1 for a 6 MV beam) for a 28 keV effective energy beam. This is compared to an over response of 9.2 at the same energy for EDR-2 film. The dose response at the maximum absorption peak was found to be approximately 3.8 and 1.9 times more sensitive than MD-55-1 and MD-55-2 films, respectively. At the absorption peak yielding the maximum optical density change, HS was found to be approximately 0.2 to 0.25 times the sensitivity of EDR-2.

Purpose: Silicon diode arrays are commonly implemented in radiation therapy quality assurance applications as they have a number of advantages including: real time operation (compared to the film) and high spatial resolution, large dynamic range and small size (compared to ionizing chambers). Most diode arrays have detector pitch that is too coarse for routine use in small field applications. The goal of this work is to characterize the two-dimensional monolithic silicon diode array named “MagicPlate-512” (MP512) designed for QA in stereotactic body radiation therapy (SBRT) and stereotactic radio surgery (SRS). Methods: MP512 is a silicon monolithic detector manufactured on ap-type substrate. An array contains of 512 pixels with size 0.5 × 0.5 mm{sup 2} and pitch 2 mm with an overall dimension of 52 × 52 mm{sup 2}. The MP512 monolithic detector is wire bonded on a printed circuit board 0.5 mm thick and covered by a thin layer of raisin to preserve the silicon detector from moisture and chemical contamination and to protect the bonding wires. Characterization of the silicon monolithic diode array response was performed, and included pixels response uniformity, dose linearity, percent depth dose, output factor, and beam profiling for beam sizes relevant to SBRT and SRSmore » and depth dose response in comparison with ionization chamber. Results: MP512 shows a good dose linearity (R{sup 2} = 0.998) and repeatability within 0.2%. The measured depth dose response for field size of 10 × 10 cm{sup 2} agreed to within 1.3%, when compared to a CC13 ionization chamber for depths in PMMA up to 30 cm. The output factor of a 6 MV Varian 2100EX medical linac beam measured by MP512 at the isocenter agrees to within 2% when compared to PTW diamond, Scanditronix point EDD-2 diode and MOSkin detectors for field sizes down to 1 × 1 cm{sup 2}. An over response of 4% was observed for square beam size smaller than 1 cm when compared to EBT3 films, while the beam profiles (FWHM) of MP512 match to within 2% the data measured by radiochromic film. Conclusions: The response of the 2D detector array, MP512, has been evaluated. The properties of the array demonstrated suitability for use as in phantom dosimeter for QA in SRS and SBRT. Although MP512 matches film measurements down to 1 × 1 cm{sup 2} well, it showed a discrepancy of 4% in the determination of output factors of beams smaller than 0.5 × 0.5 cm{sup 2} due to the field perturbation generated by the large amount of silicon surrounding the central diode. MP512 is highly capable of measuring beam size (FWHM) and has a discrepancy of less than 1.3% when compared to EBT3 film. A reduction in the detector pitch to less than 2 mm would improve the penumbra reconstruction accuracy at the cost readout electronics complexity.« less

The paper presents a comparative study performed by the Centre of Medical Radiation Physics (CMRP) on three multichannel Data Acquisition Systems (DAQ) based on different analogue front-ends to suit a wide range of radiotherapy applications. The three front-ends are: a charge-to-frequency converter developed by INFN Torino, an electrometer and a charge-to-digital converter (both commercial devices from Texas Instruments). For the first two (named DAQ A and B), the CMRP has designed the read-out systems whilst the third one (DAQ C) comes with its own evaluation board. For the purpose of the characterization DAQ A and DAQ B have been equipped with 128 channels while DAQ C has 256 channels. In terms of performances, the DAQs show good linearity over all the dynamic range. Each one has a different range of sensitivity ranging from less than 1 pC up to 13 nC, which makes the three front-ends complementary and suitable for use with different radiation detectors for different radiotherapy applications, or in a mixed solution which can house different front-ends.

Purpose: A potential side effect of inline MRI-linac systems is electron contamination focusing causing a high skin dose. In this work, the authors reexamine this prediction for an open bore 1 T MRI system being constructed for the Australian MRI-Linac Program. The efficiency of an electron contamination deflector (ECD) in purging electron contamination from the linac head is modeled, as well as the impact of a helium gas region between the deflector and phantom surface for lowering the amount of air-generated contamination. Methods: Magnetic modeling of the 1 T MRI was used to generate 3D magnetic field maps both with and without the presence of an ECD located immediately below the MLC’s. Forty-seven different ECD designs were modeled and for each the magnetic field map was imported into Geant4 Monte Carlo simulations including the linac head, ECD, and a 30 × 30 × 30 cm3 water phantom located at isocenter. For the first generation system, the x-ray source to isocenter distance (SID) will be 160 cm, resulting in an 81.2 cm long air gap from the base of the ECD to the phantom surface. The first 71.2 cm was modeled as air or helium gas, with the latter encased between two windows of 50 μm thick high density polyethlyene. 2D skin doses (at 70 μm depth) were calculated across the phantom surface at 1 × 1 mm2 resolution for 6 MV beams of field size of 5 × 5, 10 × 10, and 20 × 20 cm2. Results: The skin dose was predicted to be of similar magnitude as the generic systems modeled in previous work, 230% to 1400% of for 5 × 5 to 20 × 20 cm2, respectively. Inclusion of the ECD introduced a nonuniformity to the MRI imaging field that ranged from ∼20 to ∼140 ppm while the net force acting on the ECD ranged from ∼151 N to ∼1773 N. Various ECD designs were 100% efficient at purging the electron contamination into the ECD magnet banks; however, a small percentage were scattered back into the beam and continued to the phantom surface. Replacing a large portion of the extended air-column between the ECD and phantom surface with helium gas is a key element as it significantly minimized the air-generated contamination. When using an optimal ECD and helium gas region, the 70 μm skin dose is predicted to increase moderately inside a small hot spot over that of the case with no magnetic field present for the jaw defined square beams examined here. These increases include from 12% to 40% of for 5 × 5 cm2, 18% to 55% of for 10 × 10 cm2, and from 23% to 65% of for 20 × 20 cm2. D max D max D max D max Conclusions: Coupling an efficient ECD and helium gas region below the MLCs in the 160 cm isocenter MRI-linac system is predicted to ameliorate the impact electron contamination focusing has on skin dose increases. An ECD is practical as its impact on the MRI imaging distortion is correctable, and the mechanical forces acting on it manageable from an engineering point of view.

Cartesian co-ordinates, traditionally used for radiotherapy margins, calculated at 6 points, may not adequately represent changes in inter-observer contour variation as necessary to define a delineation margin. As a first step, this study compared the standard deviation (SD) in contour delineation using Polar and Cartesian co-ordinates for whole breast. Whole breast Clinical Target Volumes (CTV) were delineated by eight observers for 9 patients. The SD of contour position was determined for Polar co-ordinates at 1° increments for 5 slices and averaged across all patients. The mean centre of mass (COM) was used as the origin for the right breast, for the left the COM was shifted 1cm superiorly to avoid clipping. The SD was determined for Cartesian co-ordinates for medial-lateral and anterior-posterior positions. At slice Z=0cm considering Polar co-ordinates, the SD peaked medially reaching 3.55cm at 15° for the right breast, and 1.44cm at 171° for the left. The SD of the remaining slices maintained a similar distribution, with variation in the peak occurring within 10° of the Z=0cm positions. By comparison, for Cartesian co-ordinates at slice Z=0cm, the largest SD in the medial-lateral and anterior-posterior directions was 0.54/0.57cm and 1.03/0.67cm respectively for right/left breasts. The SD for inter-observer variation for whole breast varies with anatomical position. The maximum SD determined with Polar co-ordinates was greater than with Cartesian co- ordinates. A delineation margin may thus need to vary with angle over the entire structure and Cartesian co-ordinates may not be the best approach for margin determination for whole breast.

Sufficient skin dose needs to be delivered by a radiotherapy chest wall treatment regimen to ensure the probability of a near surface tumor recurrence is minimized. To simulate a chest wall treatment a hemicylindrical solid water phantom of 7.5 cm radius was irradiated with 6 MV x-rays using 20x20 cm2 and 10x20 cm2 fields at 100 cm source surface distance (SSD) to the base of the phantom. A surface dose profile was obtained from 0 to 180 degrees, in 10 degrees increments around the circumference of the phantom. Dosimetry results obtained from radiochromic film (effective depth of 0.17 mm) were used in the investigation, the superficial doses were found to be 28% (of Dmax) at the 0 degrees beam entry position and 58% at the 90 degrees oblique beam position. Superficial dose results were also obtained using extra thin thermoluminescent dosimeters (TLD) (effective depth 0.14 mm) of 30% at 0 degrees, 57% at 90 degrees, and a metal oxide semiconductor field effect transistor (MOSFET) detector (effective depth 0.5 mm) of 43% at 0 degrees, 62% at 90 degrees. Because the differences in measured superficial doses were significant and beyond those related to experimental error, these differences are assumed to be mostly attributable to the effective depth of measurement of each detector. We numerically simulated a bolus on/bolus off technique and found we could increase the coverage to the skin. Using an alternate "bolus on," "bolus off" regimen, the skin would receive 36.8 Gy at 0 degrees incidence and 46.4 Gy at 90 degrees incidence for a prescribed midpoint dose of 50 Gy. From this work it is evident that, as the circumference of the phantom is traversed the SSD increases and hence there is an inverse square fluence fall-off, this is more than offset by the increase in skin dose due to surface curvature to a plateau at about 90 degrees. Beyond this angle it is assumed that beam attenuation through the phantom and inverse square fall-off is causing the surface dose to reduce.

It has been shown that radiochromic film is an ideal dosimeter for assessment and verification of delivered dose to irradiated blood products. Using a parallel opposing two-field technique on a medical linear accelerator, blood is irradiated to diminish the risk of transfusion-associated graft vs. host disease (TA-GVHD). The blood products are irradiated in a Perspex blood box to an applied dose of 29.5-31.7 Gy. Verification of applied dose has been performed with thimble ionization chambers and radiochromic film. Radiochromic film results have matched absorbed dose measurements from ionization chambers at all sites within the 'active' treatment volume within +/-6% for a 95% confidence limit. Using a sample of 100 in-vitro measurements, radiochromic film has measured the average applied dose to blood products to be 30.95+/-2.6 Gy for two standard deviations. Like currently available 'irradiated' film labels, the radiochromic film also serves as a visible reminder that the blood products have been irradiated.

Radiographic film is an established practical tool used in the measurement of the dose distribution for radiotherapy purposes. The accuracy and reproducibility of film optical density as an indicator of dose has been associated with several factors including photon energy, processing conditions and film plane orientation. Few studies have investigated the factors causing variability in film dosimetry, due to the difficulty of separating the individual contributions. The effect that a sheet of radiographic film in a water phantom has on its response to a 6 MV photon and a cobalt-60 teletherapy beam, when orientated perpendicular and parallel to the beam central axis, is reported. Monte Carlo generated spectra were used to calculate collision kerma (Kcoll) for water and film elements. Measured and calculated results indicate a potential over-response at 25 cm depth of the order of 14 +/- 2.4% and 18 +/- 6.0% respectively for 6 MV photons and 15 +/- 3.4% and 32 +/- 4.5% respectively for a cobalt beam. For film exposed parallel as compared to perpendicular to the central axis of the beam, the calculated results suggest an explanation in terms of the predominantly forward directed secondary electrons for the measured difference in film response at depth. It is proposed that the difference in response of the parallel as compared to perpendicular exposed film be due to the predominantly 'upstream' photon interactions giving rise to energy deposition in film. The simulations indicate that the variation with depth of relative energy imparted in film and water elements correlates with the observed variation in film response with depth.

This study was undertaken to investigate the possibility of increasing the dose per fraction or increasing the number of fractions to account for treatment delays occurring during radiotheraphy treatments for highly proliferative tumours. The linear quadratic model with time was used to determine the difference in biological effective dose (BED) for the original schedule and the schedule including a treatment delay. Tables of extra fractions and extra dose per fraction required to account for a number of possible delays have been determined. It has been shown that for tumours with very short potential doubling times it is best to deliver the extra dose as an increase in dose per fraction rather than an increase in the number of fractions, while for tumours with moderately short potential doubling times (above 7 days) the reverse is true. The equivalent uninterrupted schedules, which would have delivered the same effects to the tumour, have also been determined.

Both read-out light sources and ambient light sources can produce a marked effect on coloration of radiochromic film. Fluorescent, helium neon laser, light emitting diode (LED) and incandescent read-out light sources produce an equivalent dose coloration of 660 cGy h(-1), 4.3 cGy h(-1), 1.7 cGy h(-1) and 2.6 cGy h(-1) respectively. Direct sunlight, fluorescent light and incandescent ambient light produce an equivalent dose coloration of 30 cGy h(-1), 18 cGy h(-1) and 0 cGy h(-1) respectively. Continuously on, fluorescent light sources should not be used for film optical density evaluation and minimal exposure to any light source will increase the accuracy of results.

In this study a simple method using standard flood-field corrected Electronic Portal Imaging Device (EPID) images for routine Intensity Modulated Radiation Therapy (IMRT) Quality Assurance (QA) was investigated. The EPID QA system was designed and tested on a Siemens Oncor Impression linear accelerator with an OptiVue 1000ST EPID panel (Siemens Medical Solutions USA, Inc, USA) and an Elekta Axesse linear accelerator with an iViewGT EPID (Elekta AB, Sweden) for 6 and 10 MV IMRT fields with Step-and-Shoot and dynamic-MLC delivery. Two different planning systems were used for patient IMRT field generation for comparison with the measured EPID fluences. All measured IMRT plans had >95% agreement to the planning fluences (using 3 cGy / 3 mm Gamma Criteria) and were comparable to the pass-rates calculated using a 2-D diode array dosimeter.

Purpose: To quantify the magnitude and distribution of system related geometric distortion present in MR images and determine the clinical radiotherapy sites for which this distortion is likely to have most impact. Methods: A commercially developed phantom conforming to AAPM standards for testing uniformity and linearity was imaged on 4 different MRI scanners. Similar T1 and T2 weighted clinical protocols were assessed. Each sequence and scanner had varying protocol parameters. Additionally, the direct impact of altering the receiver bandwidth (7.69 – 76.8kHz) on distortion and signal to noise was conducted on a standard spin echo sequence. The position of each phantom grid point was determined using in‐house MATLAB code, then compared to the known position as verified by a CT scan of the phantom to obtain the deformation matrix. Positions of anatomical sites (breast, lung, cervix, prostate and H&N) were determined by analysing radiotherapy contoured data for 10 patients per site to establish the range of distortions occurring within these regions. Results: Distortions were observed to range from 0mm – 8.9mm in magnitude. Maximum distortion values measured across one of the similar clinical spin echo sequences tested varied from 2.75mm – 8.94mm. Similarly, comparison of a gradient echo sequence showed maximum distortions varied from 2.56mm – 3.33mm. Increasing the bandwidth reduced the maximum distortion present in the image by 3.49mm, at the cost of SNR. While distortion magnitudes and distributions varied, in most sequences the distortions began exceeding 2mm beyond a radial distance of 100mm from isocentre, where the oral cavity, lungs and breast lie. For regions where the prostate, cervix and brain stem lie, distortions were below 2mm. Conclusion: This work displays the variations in distortions present in MR images for different sequences and scanners. Distortions are of greatest magnitude in areas where the lung and breast would be located for radiotherapy purposes.

Purpose: Consistency of radiotherapy contours is required to ensure consistency of treatment and for this reason many studies have been undertaken and are expected in the future comparing contours of multiple observers or systems. This study was undertaken to determine the minimum uncertainty achievable when undertaking this type of investigation including multiple centres and treatment planning systems. Methods: A Computed Tomography (CT) scan was taken of a commercially available uniformity Phantom. This dataset was then imported into various contouring software programs including Pinnacle, Xio and Focal at the same institution and variations at different institutions. Contours of the perimeter of the phantom and a detailed cylinder inside the phantom were contoured using the same observer at provided window levels. The perimeter of the phantom was auto‐contoured using auto‐threshold. The inside circle was contoured manually. Contours were then exported from the treatment planning systems and into CERR for analysis. Results: A comparison of the phantom perimeter from Focal and Pinnacle at a single institution demonstrated a Concordance Index (CI) of 0.98, while the manually contoured cylinder has a CI of 0.77. When comparing between institutions the CI ranged from 0.75–0.85 for the cylinder. Variation in the phantom perimeter contours was mainly in the Z direction with 2 slices (0.4cm) not being contoured in Focal compared to Pinnacle. Maximum variation in the X and Y direction for the phantom perimeter was 0.098cm. The centre of mass of all phantom perimeter contours were within 0.10cm, with the largest variance between institutions occurring in the anterior‐posterior direction. Conclusion: The variation between auto‐contouring and manually contouring a high contrast object for different treatment planning systems has been established. As expected manually contouring produces greater variation than auto‐threshold contours between different treatment planning system. Funding from Cancer Australia and The National Breast Cancer Foundation, Project grant 1033237

Purpose: The conformity of the achieved dose distribution to the treatment plan strongly correlates with the accuracy of seed implantation in a prostate brachytherapy treatment procedure. Incorrect seed placement leads to both short and long term complications, including urethral and rectal toxicity. The authors present BrachyView, a novel concept of a fast intraoperative treatment planning system, to provide real-time seed placement information based on in-body gamma camera data. BrachyView combines the high spatial resolution of a pixellated silicon detector (Medipix2) with the volumetric information acquired by a transrectal ultrasound (TRUS). The two systems will be embedded in the same probe so as to provide anatomically correct seed positions for intraoperative planning and postimplant dosimetry. Dosimetric calculations are based on the TG-43 method using the real position of the seeds. The purpose of this paper is to demonstrate the feasibility of BrachyView using the Medipix2 pixel detector and a pinhole collimator to reconstruct the real-time 3D position of low dose-rate brachytherapy seeds in a phantom. Methods: BrachyView incorporates three Medipix2 detectors coupled to a multipinhole collimator. Three-dimensionally triangulated seed positions from multiple planar images are used to determine the seed placement in a PMMA prostate phantom in real time. MATLAB codesmore » were used to test the reconstruction method and to optimize the device geometry. Results: The results presented in this paper show a 3D position reconstruction accuracy of the seed in the range of 0.5-3 mm for a 10-60 mm seed-to-detector distance interval (Z direction), respectively. The BrachyView system also demonstrates a spatial resolution of 0.25 mm in the XY plane for sources at 10 mm distance from Medipix2 detector plane, comparable to the theoretical value calculated for an equivalent gamma camera arrangement. The authors successfully demonstrated the capability of BrachyView for real-time imaging (using a 3 s data acquisition time) of different brachytherapy seed configurations (with an activity of 0.05 U) throughout a 60 Multiplication-Sign 60 Multiplication-Sign 60 mm{sup 3} Perspex prostate phantom. Conclusions: The newly developed miniature gamma camera component of BrachyView, with its high spatial resolution and real time capability, allows accurate 3D localization of seeds in a prostate phantom. Combination of the gamma camera with TRUS in a single probe will complete the BrachyView system.« less

Purpose: To determine the difference between expected and measured dose for patients prescribed a mantle treatment for Hodgkin's disease and estimate the range of dose at critical sites and between different treatment centres. Methods and materials: Twenty three radiotherapy centres were surveyed with regard to the accuracy of dose delivery to a custom-built upper-torso phantom. Thermoluminescent dosimeters were used to monitor the delivered dose at sites such as mid-plane, spinal cord, neck, axilla and lung. Results: The intended dose to the phantom at each centre was 1 Gy to central axis mid-plane. Of the centres surveyed, the median measured dose to this region was 0.96 Gy with a minimum of 0.92 and a maximum of 1.00 Gy. Median dose to the axilla region was low (0.90 Gy) whereas median dose to the blocked lung and neck region were higher than expected, 0.18 Gy and 1.25 Gy, respectively. The 95% confidence interval on the reported relative dose using the 4000 thermoluminescence dosimetry readings in this study was ±1.5%. Conclusion: In this controlled experiment, using conventional methods to calculate dose, there was a surprising variability in the dose delivered at the central axis mid-plane position. This was traced to lack of uniformity in the use of equivalent squares to calculate the output factor. The measured doses to axilla and lung are explained by photon and electron scattering effects. Centres, where dose compensation was included, had a superior dose homogeneity in the neck. The off-axis dose calculations depend on computer planning software but the magnitude of these differences is secondary to that of central axis mid-point dose differences. Improved consistency of dose calculation techniques between centres would enable more reliable dose response evaluation from multicentre clinical studies of Hodgkin's disease.

Neodymium Iron Boron (NdFeB) rare earth permanent magnets have unique properties that enable them to fit easily onto the accessory mount of a clinical linear accelerator to partially sweep away electron contamination produced by the treatment head and block trays and thus increase skin sparing. Using such magnets the central axis entrance surface dose has been reduced by 11% for a 20 x 30 cm field size from 32% to 21% of maximum dose by the magnetic device. A reduction of 14% from 32% to 18% was seen for a 20 x 20 cm field size with a 6 mm perspex block tray positioned above the magnet. The magnetic device is light weight and thus clinically usable.

Radiochromic film is shown to be a useful surface-dose detector for radiotherapy x-ray beams. Central-axis percentage surface-dose results as measured by Gafchromic film for a 6 MVp x-ray beam produced by a Varian 2100C Linac at 100 cm SSD are 16%, 25%, 35%, 41% for 10, 20, 30 and 40 cm square field sizes, respectively. Using a simple, uniform light source and a CCD camera connected to an image analysis system, quantitative 3D surface doses are accurately attainable in real time as either numerical data, a black-and-white image or a colour-enhanced image.

Purpose/Objective(s): As radiation therapy becomes more conformal and the delivery technology becomes more complex, in-vivo dosimetry is becoming more necessary but at the same time more technically challenging. This work investigates the feasibility of a novel integrated in-vivo imaging and dosimetry device to meet the current and future requirements of radiation therapy treatment verification whereby imaging and dosimetry could be achieved concurrently. Materials/Methods: The dose response of a commercial 2-D ion-chamber array, in two configurations was investigated; (1) the experimental configuration with the ion-chamber array positioned directly under an electronic portal imaging device (EPID), (2) the standard dosimetry configuration with the ion-chamber array and 5cm solid water build-up. All measurements were conducted with the ion chamber array at isocentric distance and no additional objects in the beam. The measured dose was compared with the calculated dose from a treatment planning system (TPS). For the experimental configuration (1) the ion chamber array was positioned on the treatment couch with the EPID sitting directly on top of it. An additional 3.2 cm solid water was added to fill the air gap between the array and EPID. All the experiments were conducted at zero gantry angle. The central axis dose per Monitor unit and dose profiles for a 10x10 cm open field was measured in both configurations. The water equivalent depth of the experimental configuration (1) was determined by comparison with the TPS calculated dose. This depth was subsequently used for TPS dose calculations for further open beams, IMRT test patterns, and clinical IMRT beams. The field size factor and off-axis dose response measured with configuration (1) for open fields was compared with TPS data. The gamma analysis was performed with 2% and 3 mm criteria. Agreement between ion-chamber array and TPS was compared for both configurations. Results: The water equivalent depth of the ion chamber array in configuration (1) was 8.2 cm. The maximum dose difference between measured dose profiles with configuration (1) and TPS was 3.5% for 20x20 cm field size. The field size output factor with configuration (1) agreed within 2% with TPS. For a given set of gamma criteria the percentage of gamma pass agreed to with 0.5% for all the measurements. Conclusion: An ion chamber array integrated with an EPID maintains accurate water equivalent dose response. This configuration has the potential to provide a new technique of accurate 2D in vivo dosimetry. Further work is underway to assess the impact of the ion chamber array on imaging, and the performance of the device with an object in the beam to more closely simulate the clinical setting. Author Disclosure: S. Deshpande: None. L. Holloway: None. P. Metcalfe: None. P. Vial: None.

Purpose: To examine via Monte Carlo simulation the dosimetric operation of an EPID inside a bi‐directional MRI‐Linac system. The proposed system is a 1 T MRI which supports both perpendicular (transverse) and inline (longitudinal) 6 MV MRIgRT. Methods: BEAMnrc Monte Carlo simulations were used to model a Varian 6 MV linac. Phase space files were produced below the MLCs and were inputs to Geant4 Monte Carlo simulations. These consist of transporting the radiation beam through a phantom and then on to a detailed model of a Varian EPID. The Geant4 simulations take place inside magnetic field maps matched to Finite Element Modeling of the MRI‐Linac system (COMSOL). Hence the EPID operation is simulated in the same conditions as a real MRI‐linac system. Various EPID SSD's were simulated in both inline and perpendicular field MRI‐linac systems. Dose to the EPID layers were taken for each field and compared with zero magnetic field. Results: The inclusion of the magnetic fringe field had little impact in the inline direction, however in the perpendicular direction the electron‐return effect (ERE) caused secondary electrons to return to the phosphor scoring level. This induced a lateral blur and overdosing of the images that was related directly to the SSD or strength of the fringe field in the transverse direction (5% at 0.25 T, 10% at 1 T). Including 1 cm of water material below the phosphor layer significantly reduced the ERE and improved the dose image quality. Conclusion: Novel Monte Carlo simulations of the operation of an EPID working inside a bidirectional field MRI‐linac have been performed. The main effect which induced an image or dose artifact was the ERE in the perpendicular field arrangement. This can however be almost eliminated by the use of an exit side bolus on the EPID.

Purpose: To perform a comparative study assessing potential benefits of endorectal balloons (ERB) in post‐prostatectomy patients Methods: Ten retrospective post‐prostatectomy patients without ERB and ten prospective patients with the ERB in situ were treated with IMRT using kilo voltage cone beam computed tomography (kVCBCT) image guidance (IGRT). A kVCBCT dataset from each treatment week was registered to the planning dataset, recontoured and the original plan recalculated on the kvCBCTs to recreate actual anatomical conditions during treatment. The imaging, structure and dose data were imported into Matlab. In house applications analysed geometric variation in the two groups and calculated cumulative equivalent uniform dose (EUD). Results:Analysis of the kVCBCT datasets showed that the ERB improved target volume stabilization from millimetre to sub‐millimetre reproducibility. The variation in bladder location in the non ERB cohort was 1.5 ± 0.9 cm compared to 0.6 ± 0.4 in the ERB group. While the superior rectum location variation was 1.1 ± 0.8 cm and 0.3 ± 0.3 cm for non ERB and ERB respectively. The mean difference in cumulative EUD and planned EUD for the clinical target volume (CTV) was 0.6 ± 0.04 Gy for non ERB and 0.4 ±0.1 Gy for ERB. The largest normal tissue differences were seen for the bladder with an increase of 4.4 ± 0.1 Gy for non ERB and 1.2 ± 0.02 Gy for ERB in the cumulative compared with the planned EUD. Differences in delivered superior rectal EUD was improved in the ERB cohort with a difference of−0.2 ± 0.04 Gy verse 3.0 ± 0.3 Gy in the non ERB group. Conclusions: The ERB cohort was found to have more stable and reproducible anatomy when compared to the non ERB cohort. Planned and cumulative EUD differences were greatest for the bladder and rectum.

Purpose: This work investigates the ratio of the two source components of the undesirable skin dose increases seen in transverse field MRI‐Linac systems: the Lorentz‐force perturbation (LFP) and the electron return effect (ERE). Dose planning systems using kernels generated in magnetic fields will account for the LFP, however not the ERE. This work indicates the inaccuracies in not modeling the ERE in conventional planning. Methods: Geant4 Monte Carlo simulations of a Varian 2100C 6MV photon beam in the presence of transverse magnetic fields are performed. By using tracking options the dose deposited by the ERE can be compared with that delivered by both the LFP. The effect of surface angle and field size is also examined on a phantom of 30×30×20cm3. High resolution scoring voxels are used to extract the 70 micron skin dose values. Results: On the phantom entry side skin dose changes are clinically significant at large positive angles. The ERE causes about 50% of this depending on magnetic field strength. At negative angles the ERE plays a minimal role in the skin dose increases seen. On the phantom exit side the skin dose increase is the greatest at negative surface angles, and furthermore, this is almost entirely due to the ERE electrons. For positive surface angles the ERE only gives rise to a small fraction of the exit skin dose increases. Conclusion: The skin dose increases seen in 6MV transverse field MRI‐Linac systems are significant at positive angles on the entry side, and at negative angles on the exit side. The ERE can cause a large proportion of these changes, which will not be inherently modeled correctly inside a typical dose planning engine. These results suggest that improvements to current planning systems are needed if accurate estimates of the skin dose changes in MRI‐guided‐radiotherapy are required.

Purpose: To investigate the behavior of the imagingcontrol software on a Siemens EPID and the resulting impact on EPID based IMRTdosimetry Method and Materials:Four ‘simple’ and four ‘clinical’ IMRT fields were created. Each segment of the ‘simple’ field had a constant field size of 10 × 10 cm2. The ‘simple’ fields had 2, 4, 5, and 10 segments with 10, 5, 4, and 2 MU per segment respectively. The ‘clinical’ fields all consisted of 20 segments with constant MU per segment of 2, 4, 5, and 10 MU. The fields were measured simultaneously with a Siemens EPID and ImRT MatriXX ionisation chamber array (iba) by placing the MatriXX beneath EPID. The integrated EPID dose response was determined from the frame average pixel value multiplied by the reported number of frames. Reproducibility of central axis response, beam profiles, and pixel‐by‐pixel response (2D) was compared from 10 repeat measurements. Results: The percentage standard deviation of the integrated EPID response for the ‘simple’ case was within 1% for the 2 and 10 MU per segment fields but more than 10% for the 4 and 5 MU per segment fields. EPID crossplane profiles also yielded large variations in amplitude and shape for the 4 and 5 MU per segment cases only. No such variations were observed by concurrent MatriXX measurements. A similar but less obvious trend was observed for the ‘clinical’ case. In both cases the number of frames per segment reported for the 2 MU and 10 MU per segment fields remained constant. The number of frames reported for the 4 MU and 5 MU per segment fields varied and appeared unreliable. Conclusions: EPIDimages of IMRT fields displayed large variations in reproducibility depending on the MU per segment. Poor reproducibility correlated with inconsistent reporting of the number of frames. Cancer Institute NSW Research Equipment Grant 10/REG/1‐20

Purpose: This work uses Monte Carlo simulations to model an unexpected dosimetry effect of electron contamination in Longitudinal Field MRI‐Linac systems (LF‐MRIGRT). These systems are under construction due to their dosimetric advantages over transverse‐field systems. However this work has yielded a potential major dosimetric problem and aims to investigate its scope. Methods: Geant4 Monte Carlo simulations of various generic longitudinal‐field MRI‐Linac systems are performed. The fringe‐field of the MRI system is included as it collects and propagates electron contamination from the Linac towards the patient. Field size, isocentre distance, and fringe‐ field style are all varied and examined. The Linac modeled is a Varian 2100c 6MV beam, and dose is scored in a 30×30×20cm3 phantom. Results: The longitudinal component of the MRI‐system fringe field acts to trap electron contamination arising from the Linaccollimation system. This becomes focused and cumulates as it travels towards the patient level. As a result a massive concentration of electron contamination is present near CAX at the patient level, leading to entry skin dose hot spots of > 500% of dose at dmax (15 mm). Even if the Linaccollimation system is fully magnetically shielded there is resultant lack of lateral spread (or dilution) of contamination by the main MRI‐field region. This still leads to skin dose hot spots approaching Dmax. Conclusion: Monte Carlo modeling predicts that electron contamination from the Linac in LF‐MRIGRT may be trapped by the MRI system and cause considerable patient entry skin dose hot spots near CAX, of the order of > 500% of Dmax. It is highly recommended that accurate modeling of the MRI fringe‐field and MRI‐Linac system be performed to ascertain this risk in potential TF‐MRI‐Linac systems.

Data on the dose distribution in the penumbral region of megavoltage x rays are of importance for most radiotherapy planning systems. For medical linear accelerators the distance between the points representing 20% and 80% of the central axis dose (20/80) is typically only a few millimeters. To achieve good spatial resolution a radiation detector with small physical size has to be used for penumbra measurements. The penumbra in a 6-MV therapeutic x-ray beam was investigated for field sizes of 10 x 10 cm2 and 20 x 20 cm2 at depth of maximum dose (dmax = 1.5 cm) and at 10-cm depth in a solid water phantom. In addition, the field edge of an independent jaw driven to the center of the axis of the primary collimator was investigated. LiF thermoluminescent (TL) ribbons and rods were used embedded in solid water in different geometries resulting in relative detector sizes of 1, 3.1, and 6 mm. A forming function based on an inverse square root function was used to fit the experimental penumbra measurements. For the asymmetric field an amendment to the function is proposed to give a better fit for the experimental data. From the penumbra measured with the three different detector sizes, a penumbra can be extrapolated for an infinitesimal small detector. The extrapolated penumbral width (20/80) was found to be 2.3, 3.2, and 2.7 mm at dmax for the 10 x 10-cm2 symmetric, 10 x 10-cm2 asymmetric, and 20 x 20-cm2 field sizes, respectively. The 20/80 values at 10-cm depth in the solid water phantom for the same radiation fields were 4.2, 4.3, and 4.1 mm, respectively.

Surface dose measurements in therapeutic x-ray beams are of importance in determining the dose to the skin of patients undergoing radiotherapy. Measurements were performed in the 6-MV beam of a medical linear accelerator with LiF thermoluminescence dosimeters (TLD) using a solid water phantom. TLD chips (surface area 3.17 x 3.17 cm2) of three different thicknesses (0.230, 0.099, and 0.038 g/cm2) were used to extrapolate dose readings to an infinitesimally thin layer of LiF. This surface dose was measured for field sizes ranging from 1 x 1 cm2 to 40 x 40 cm2. The surface dose relative to maximum dose was found to be 10.0% for a field size of 5 x 5 cm2, 16.3% for 10 x 10 cm2, and 26.9% for 20 x 20 cm2. Using a 6-mm Perspex block tray in the beam increased the surface dose in these fields to 10.7%, 17.7%, and 34.2% respectively. Due to the small size of the TLD chips, TLD extrapolation is applicable also for intracavity and exit dose determinations. The technique used for in vivo dosimetry could provide clinicians information about the build up of dose up to 1-mm depth in addition to an extrapolated surface dose measurement.

Purpose: The main focus of this work is to continue investigations into the Monte Carlo predicted skindoses seen in MRI‐guided radiotherapy (MRIgRT). In particular we aim to characterise the 70 micron skindoses. The effect of surface orientation on both the entry and exit sides is also investigated, as well as the use of exit bolus. Materials and Methods: High resolution Geant4 Monte Carlo simulations of a water phantom exposed to a 6 MV X‐ray beam (Varian 2100C) have been performed. Transverse magnetic fields of strengths between 0 and 3 T have been applied to a 30×30×20 cm3 phantom. This phantom is also altered to have variable entry and exit surfaces with respect to the beam central axis and they range from −75 to +75 deg. The exit bolus simulated is a 1 cm thick (water equivalent) slab located on the beam exit side. Results: On the entry side significant skindoses at the beam central axis are reported for large positive surface angles and strong magnetic fields. Over the entry surface angle range of −30 to −45 deg the entry skindose is comparable with zero magnetic fieldskindose, regardless of magnetic field strength and field size. On the exit side moderate skindose increases are expected except for large positve angles at the higher magnetic fields. For exit bolus of 1 cm thick the central axis exit skindose becomes an almost consistant value at around 35–38% of Dmax. This is due to the almost complete absorption of the ERE electrons by the bolus. Conclusion: There are many combinations in MRIgRT where skindose may become excessive. Results from this study can be used to determine arrangements where skindose is comparable to B = 0 T.

Review(s) of: David R. Wigg. Applied Radiobiology: Continuous Irradiation and Brachytherapy, Medical Physics Publishing, Madison, WI, USA, ISBN 9781930524392, CD Rom Equations ISBN 9781930524453, software developed by BOLD Enterprises PTY LTD.

Total and primary polyenergetic dose spread arrays (PDSA) have been generated for a high energy 10 MV radiotherapy photon beam using the electron gamma shower (EGS) Monte Carlo code. By considering the attenuation of fluence per energy interval, PDSA have been produced at radiological depths of 0 cm (the surface PDSA) and 40 cm (the beam hardened PDSA). By comparing primary PDSA produced at these different depths, the effect of beam hardening on the PDSA has been quantified. Calculations show that the mean electron range due to the surface primary PDSA is 6.67 mm and the mean electron range of the beam hardened primary PDSA is 8.24 mm. In comparison, a 3 MeV primary monoenergetic dose spread array (MDSA) has a much smaller mean electron range of 4.81 mm. A radiotherapy x-ray beam computation method is introduced which involves a single superposition of the surface generated PDSA or beam hardened PDSA with a polyenergetic TERMA. The mean percentage difference between depth-dose curves obtained using super-position of surface and beam hardened PDSA is only 0.1%. The mean percentage difference from experimental data for these superposition curves is 2.8% down to 40 cm in a homogeneous phantom. The superposition process is shown to be forgiving to spectral differences when calculating the PDSA, but sensitive to the incident photon energy spectrum used to calculate the TERMA.

Purpose: To examine and diminish the IMRT matchline effect caused by curved leaf end design of MLC using a Monte Carlo simulation.Method and Materials: The procedure is divided into four phases. The first is validating the Monte Carlo simulation using BEAMnrc, DOSXYZnrc, and EGSnrc code to reproduce a 6 MV beam of Varian machine with 120 leaves for a conventional technique of a standard field size 10×10 cm2. The results were compared to the measurement by IC 0.01 cm3. The second phase is testing the simulation for the geometry of MLC specifically a curved leaf end design compared with Kodak XO mat V film measurement. The third phase is applying the matchline situation which generated by Pinnacle treatment planning system (step & shoot technique) into the Monte Carlo simulation. The fourth phase is applying the overlapped values to the MLC files of the Monte Carlo simulation to decrease the matchline effect of IMRT dose distribution. Results: The agreement between simulation and measurement for conventional technique was within 2% for depth dose and beam profile. The test of the geometry of MLC was done by comparison of closing leaf end of dose distribution. The result showed that the agreement was within 2%. The overlapped values were applied to the MLC leaf files and found that 0.09 cm was selected for the optimal overlapped value with some certain segment weights to compromise among matchlines. Conclusion: The Monte Carlo simulation is able to investigate the matchline effect of IMRT dose distribution. The distance of the overlapped region of MLC 0.09 cm can reduce the IMRT matchline effect. However this value may possibly depend on the weighting dose of a segmented field of IMRT technique, further investigation should be required.

SUMMARY The use of C.T. scanners in radiotherapy planning is becoming widespread. It is essential therefore that adequate assessment of the reliability of the C.T. acquired data for planning purposes is carried out. It is also important that all makes of C.T. scanner be assessed due to their varying characteristics. Dose computation inaccuracies have to be considered also if one is to get an estimate of the overall accuracy of a given radiotherapy planning technique. In general, it would appear that until the convolution techniques become commercially available the Radiotherapy computer is now the primary source of dose calculation error, due to inadquacies in commonly used inhomogeneity correction algorithms. ABSTRACT A protocol is described for assessing the accuracy of radiotherapy plans produced using C.T. derived electron density information. Essentially there are two main sources of error, those coming from various C.T. related inaccuracies and those from inaccuracies in the radiotherapy dose calculation algorithms. The various subdivisions of these two sources of inaccuracy are discussed in detail. More specifically this paper deals with the effective depth method of inhomogeneity correction, but sufficient generality is applied to the discussion to enable the same analysis to be applied to other dose calculation algorithms. It would appear that with state of the art C.T. scanning that the area which introduces the greatest errors is now radiotherapy computing accuracy. Until convolution correction techniques become commercially available this will probably continue to be the case.

Described is an experimental verification technique for cesium brachytherapy line source emission. The dose rate levels produced by a line source algorithm are compared with measurements taken around J-series stainless-steel sheathed cesium sources with a p-type semiconductor diode detector. Further study of isodensity contours is undertaken using a film dosimetry technique. The errors involved in using these techniques are assessed.

Understanding and predicting the impact of any radiotherapy treatment is critical if patients are to receive treatment with a high likelihood of eliminating the tumour and low likelihood of complications. One of the major contributing factors in determining these effects is the volume treated. This review assesses the current use and accuracy of a series of models which consider volume, building on a previous review which investigated the impact of fractionation particularly with respect to the linear quadratic model. Volume is particularly important in assessing the overall effect with respect to destroying the clonogenic cells and preventing damage to the normal tissues. Dose volume histograms are one of the simplest and most useful forms of representing volume information, however it is difficult to correlate plans based only on DVHs. For this reason various reduction schemes have been introduced and tumour control probability and normal tissues complication probability models adjusted to use this information. Many of these models have proved quite useful in the clinic although they are limited by the available radiobiological data.