Index of content:
Volume 36, Issue 6, June 2009
- Therapy Moderated Poster Session: Exhibit Hall ‐ Area 2
- Moderated Poster ‐ Brachytherapy
36(2009); http://dx.doi.org/10.1118/1.3181075View Description Hide Description
Purpose: To illustrate the utility of Depth Dose Modulation (DDM) for Ir‐192, Electronic Brachytherapy (EB) and EBDDM. Method and Materials: A numerical model was created for a miniature x‐ray source equipped with an adjustable collimator that produces a fan‐beam distribution. A general‐purpose multi‐particle transport code was used to calculate the dose distribution for multiple collimator gaps. A software program superimposed the collimated doseimages at intervals to build an overall dose distribution for a simulated micro‐stepping treatment. TG‐43 data were input to the Nucletron PLATO treatment planning system. Treatment plans were generated for breast, endometrial and lung cases; doses to target volumes and normal tissues were compared for each of the planning sources. Results: For the breast case, PTV coverage for Ir‐192, EB and EBDDM plans was 96, 100 and 94% with maximum skindoses of 100, 97 and 89% respectively. For the endometrial case, PTV coverage for the Ir‐192, EB and EBDDM plans was 99, 100 and 77% with maximum rectal doses of 140, 127 and 105% and bladder doses 115, 111 and 87% respectively. For equivalent prescription point doses in lung, with Ir‐192, EB and EBDDM, the maximum rib dose was 43, 30 and 34% while V50Lung was 11, 8 and 7% respectively. Conclusions: With intracavity breast cases, DDM is a useful tool, reducing Ir‐192 skindose by up to 11% while maintaining adequate target coverage. For endometrial treatments, however, EBDDM target coverage is inadequate; because of collimated source anisotropy, there is little dose deposition superiorly. To achieve full target coverage while maintaining significant normal tissue sparing, a treatment would require a combination of EB and EBDDM. For lung, EBDDM allows a two‐fold reduction in V50Lung. Conflict of Interest: Partial financial support provided by Xoft, Inc.
SU‐DD‐A2‐02: Monte Carlo Design of An Internally‐Shielded Peripheral Breast Brachytherapy Applicator36(2009); http://dx.doi.org/10.1118/1.3181076View Description Hide Description
Purpose: The AccuBoost® breast brachytherapy system applies HDR irradiation under mammographicimage guidance. The round‐ and D‐type applicators have been developed and are in clinical use. However, it is possible to improve upon the resultant dose distributions through introduction of an internal shield for sparing the skin. The design of this internal shield has been optimized using Monte Carlo methods to retain tumordose uniformity while minimizing skindose.Materials and Methods:Dose distributions in breast tissue from a round applicator (60 mm internal diameter) were obtained by varying the dimensions of the tungsten‐alloy internal shield. Specifically, the superior diameter and inferior diameter of the shield were varied and generally took the shape of a truncated cone. Dimensions were chosen for cone foci at tissue depths, d, of 0, 2, 5, and 10mm with field edges −1, 0, and +1mm from the source dwell plane. The MCNP5 radiation transport code was used for calculating breast dose throughout a cylindrical phantom R=15cm radius and H=8cm high. Results: Cone design 7 was determined to be the best design. This design had relatively uniform skindose distribution and relatively good retained tumordose. Average skindose was about 70% of the maximum dose, which was improved by about 27%. Average skindose was reduced by more than 44% compared to that of an applicator without a cone. The maximum reduction in skindose was 92%, which was typically outside the applicator aperture. In the mean time, average tumordose at 30 mm depth was only about 5% lower than that of applicator without a cone. Conclusions: Significant improvements in critical structure dose can be obtained through introduction of an internal shield, with design optimization using Monte Carlo methods.
36(2009); http://dx.doi.org/10.1118/1.3181077View Description Hide Description
Purpose: To develop an edema model for application in permanent prostate implant dosimetry that reflects edema spatialanisotropy and time‐resolution behaviors observed in clinical MRI data. Method and Materials: The basic manifestation of edema was represented as a trio of variations in Cartesian components of the distance, ri (t); i=1:3, from a radioactive seed to a dose calculation point; specifically . Here (r 0) i are distance components in the absence of edema, fi (t) edema time‐resolution functions, and α i quantifiers of the directional contributions to edema volume subject to the constraint . Serial MRI data from our institution for n=40 prostate implant patients is well characterized by the parameters , and the function ; t<T; i=1:3. Hence these parameters and time‐resolution function were incorporated in the model. Next, the cumulative dose from a seed to a calculation point was expressed according to the TG‐43 formalism (using the anisotropy constant) as , where λ is the radionuclide decay constant and the other variables are defined as per TG‐43. The integrand was then expanded in even powers of |r⃗(t)|, and the resulting integral in each term of the expansion evaluated to yield either a closed‐form analytic function or a convergent series. Numerical evaluation of constituent terms in the dose expansion was done in MatLab to assess functional behavior. Results: The number of terms in the expansion of D(r⃗) are few, and all are well‐behaved numerically. For those terms involving convergent series, convergence is rapid and so only a small number of terms need to be evaluated. Based on these favorable properties, MatLab‐based dose calculation software is currently being developed to investigate applications to clinical dosimetry.Conclusion: A new edema model for permanent prostate implant dosimetry incorporating spatialanisotropy has been formulated. Clinical application is pending.
SU‐DD‐A2‐04: Use of the Adjoint Analysis Based Greedy Heuristic Algorithms in Treatment Planning for LDR Brachytherapy of the Prostate and HDR Brachytherapy Using Multicatheter Breast Implant Technique36(2009); http://dx.doi.org/10.1118/1.3181078View Description Hide Description
Purpose: To demonstrate use of the adjoint analysis based Greedy Heuristic (GH) algorithms in treatment planning for LDR brachytherapy of the prostate using directional sources and multicatheter‐HDR breast implant. Method and Materials: Two adjoint analysis based GH treatment planning tools are developed; The first, for a directional LDR brachytherapy application and the second, for a HDR‐multicatheter brachytherapy application. Both the problems have an extra degree of freedom compared with conventional, binary LDR treatment planning — the seed rotational orientation in directional LDR brachytherapy and the variable dwell‐time in HDR brachytherapy. The greedy heuristic treatment planning algorithm uses adjoint‐based ROI‐sensitivity‐fields to search for the best available source type and orientation, or dwell time increment at a source‐location in steps to build either a seed‐needle‐source‐orientation distribution or dwell‐location‐dwell‐time distribution solution non‐iteratively. The GH treatment planning for LDR technique is based on dose‐distributionoptimization and that for HDR technique is based on dose‐homogeneityoptimization.Results: The treatment plans generated by the greedy heuristic algorithm using the directional sources resulted in target coverage with V100 >98% and remarkably better OAR‐sparing owing to the directional dose properties of the sources, as seen from the DVH analysis and evaluation parameters comparison on 6 prostate cases when compared with conventional LDR‐brachytherapy. The multicatheter‐HDR brachytherapy dwell‐time optimization generated treatment plan with a target coverage V100 >96%, skin‐sparing (D85=4.5%), few hot‐spots and a dose homogeneity index of 0.82. Conclusion: Greedy Heuristic algorithms coupled with greedy criterions based on ROI‐sensitivity fields of dose response and dose‐homogeneity response are utilized as efficient tools for fast and reproducible treatment planning solutions in LDR and HDR brachytherapy.
SU‐DD‐A2‐05: A Conversion Method of Low Dose Rate to Pulsed Dose Rate Intracavitary Brachytherapy Prescription for the Treatment of Cervical Carcinoma36(2009); http://dx.doi.org/10.1118/1.3181079View Description Hide Description
Purpose: To report on a retrospective clinical dosimetry analysis of the conversion of the low‐dose rate (LDR) to pulsed dose rate (PDR) intracavitary brachytherapy (ICBT). Methods and Materials: A tool was developed that converts an LDR prescription to PDR. This method incorporates the differences between LDR and PDR radial dose functions and includes perturbation factors of the different ICBT applicators as derived from MCNP Monte Carlo simulations. Absolute dwell time per PDR source position is calculated and entered into the Nucletron Plato TPS. A cohort of 67 LDR patients was randomly selected from our clinical GYN database. Plans were generated with Varian BrachyVision v6.5 for Selectron LDR sources, and with Nucletron Plato v14.1 for the PDR source.Doses were compared at the ICRU 38 patient points: Bladder, Rectum, A right and left, B right and left, 3 o'clock and 9 o'clock. A 5% difference between LDR and PDR doses at these points was considered significant. Results: In every case, the difference in dose between the LDR plan and the PDR plan was less than 5%. In addition, the differences in dose appeared to be randomly distributed about zero, implying that there was no systematic difference in the LDR and PDR dose points. Conclusions: A transfer method of our traditional LDR clinical program to PDR was validated. This insures our history with ICBT LDR experience is consistent for future ICBT patients undergoing PDR treatments.
36(2009); http://dx.doi.org/10.1118/1.3181080View Description Hide Description
Purpose: To evaluate the reliability growth in seed delivery by TJU's EUCLIDIAN, a fully automated robotic system, developed for image‐guidedbrachytherapy (IGBT). Method and Materials: Important steps in reliability growth analysis are: identification and isolation of failures, classification of failures, and trend analysis. For any one‐of‐a‐kind product, like EUCLIDIAN, reliability enhancement is accomplished through test‐fix‐test cycles of the product. Failure Mode, Effect and Criticality Analysis (FMECA) were used for collection and analysis of reliability data by identifying and categorizing the failure modes. Failures were classified according to severity. Failures that occurred in EUCLIDIAN operations were considered as Non‐Homogenous Poisson Process (NHPP), and the trend was analyzed using Laplace test. For analyzing and predicting reliability growth, commonly used and widely accepted models, Duane's model and Crow's model, were applied. Mean Time Before Failure (MTBF) was used as an important measure for assessing reliability. Results: During pre‐clinical testing, 360 seeds (in 10 cases) were deposited automatically and 6 critical failures were encountered. The majority (5 failures) of which occurred during the first three cases. The Laplace test index was −1.134 (<0), indicating significant trend in failure data, and the failure interval values were gradually becoming larger. Since the failure occurrence interval was increasing, the system reliability exhibited an increasing trend. System's failures distribution followed both the Duane's and Crow's postulations of reliability growth. MTBF was 129.4 seeds, which meant that a full brachytherapy case could be performed without any critical failure. The shape parameter for MTBF was 0.4638 (<1), suggesting positive reliability growth of the EUCLIDIAN. At 0.95 confidence, the lower and upper bounds of the MTBF were 42.8 and 612.9, respectively. Conclusion:Analyses of failure mode strongly indicated a gradual improvement in EUCLIDIAN's reliability. For better consistency in MTBF and reliability prediction, more data are being collected.
Acknowledgement: supported by NCI‐R01‐CA091763.
- Moderated Poster ‐ Treatment Planning I
SU‐EE‐A2‐01: Spatial Dose‐Volume Histogram (sDVH) — Incorporating Spatial Dose Information Back Into the DVH36(2009); http://dx.doi.org/10.1118/1.3181099View Description Hide Description
Purpose: To incorporate spatial information into dose‐volume histograms (DVH). Methods and materials: For clinical prostate and head/neck cases, initial plans were created using dose‐volume constraints for planning target volume (PTV) and organs at risk in the Eclipse treatment planning system. Dicom RT files were exported to in‐house Matlab‐based software which categorizes target dose voxels by their location within the PTV. For this study, we arbitrarily chose three regions: center, middle, and periphery. Spatial dose‐volume histograms (sDVH) were created by color‐coding the three regions within the differential DVH. Spatially optimized plans were then created using the sDVH concept to preferentially push cold spots from the target center to the periphery. Results: For both prostate and head/neck cases, the DVH and dose distributions are different for the two plans. Using DVH only, the initial plan appears more homogeneous than the spatially optimized plan, however, inspection of the dose distribution reveals that the spatially optimized plan may be clinically favorable since there is less underdosing in the center of the target. The more homogeneous initial plan has a sharper peak in the sDVH. However, it also leaves a significantly larger fraction of the center region underdosed in comparison to the spatially optimized plan (5% vs 0.5% for prostate; 11% vs 3% for head/neck). Conclusion: Treatment plan evaluation based on DVH analysis alone excludes spatial information about the dose distribution. Plans with superior DVHs might be clinically inferior once spatial dose information is considered. The sDVH is a very simple and intuitive way to incorporate spatial dose information back into the DVH format. The sDVH provides more information than DVH alone and may be more consistent with dose distribution analysis for plan ranking purposes, thus making it a useful tool for plan quality evaluation and optimization.
Conflict of interest: Supported by Varian Medical Systems.
SU‐EE‐A2‐02: Validation of a Grid‐Based Boltzmann Solver for 6 and 18 MV Photon Beams Impinging On a Heterogeneous Phantom36(2009); http://dx.doi.org/10.1118/1.3181100View Description Hide Description
Purpose: To benchmark the accuracy of a grid‐based Boltzmann solver, Acuros, against Monte Carlo for 6 and 18MV photon beams in a tissue‐bone‐lung‐tissue slab phantom. Method and Materials: Acuros was developed specifically for radiotherapydose calculations. It solves for both neutral and charged particle transport, and incorporates spatial adaptation to automatically increase spatial resolution in regions of sharp gradients such as beam penumbra and material heterogeneities. Comparisons were made between EGSnrc and Acuros for single field photon beams impinging on a 30×30×30cm3 slab phantom consisting of: tissue (0–3cm), bone (3–5cm), lung (5–12cm), and tissue (12–30cm). Calculations were performed for square fields of 2.5×2.5, 5.0×5.0, and 10×10cm2 for both 6MV and 18MV energies. Comparisons were made between Acuros and EGSnrc on 2×2×2mm3 voxels along the beam centerline (0–30cm), and on transverse lines at depths of Dmax, 10, and 20cm. To minimize the influence of Monte Carlo uncertainties, the EGSnrc calculations were run to a tight statistical uncertainty (σ≈0.1%) and fine resolution, up to 1 mm in the penumbra. Acuros results were output on a 2×2×2mm3 cubic grid encompassing the 30×30×30cm3 phantom. All comparisons were made in terms of absolute doses, Gy per incident particle. Results: For all 6 cases, maximum voxel‐wise differences between EGSnrc and Acuros along the beam centerline were less than 2% of Dmax or 1mm distance‐to‐agreement (DTA). Maximum voxel‐wise differences along the transverse lines were less than 2% of Dmax or 2mm DTA. Acuros calculation times were less than 5 minutes on a two processor AMD Opteron workstation. Conclusion: The results indicate an excellent agreement between both codes. Since Acuros calculation times are primarily dependent on the volume of the region being solved, and not on the number of beams, it is well suited for arc‐therapy modalities.
Conflict of Interest: Work funded by NIH Grant 2R44CA105806‐02.
SU‐EE‐A2‐03: Motion‐Weighted Dose‐Volume Histogram: A More Meaningful and Practical Four‐Dimensional Planning and Evaluation Method36(2009); http://dx.doi.org/10.1118/1.3181101View Description Hide Description
Purpose: 4D‐CT is frequently utilized by radiation oncologists to delineate a tumor and account for its motion. Current methods to generate the PTV do not fully utilize the 4D image data. After expanding from ITV to PTV, the 4D information is partially lost as the specific volume occupied by a tumor in any given respiratory phase is no longer apparent. Therefore, the planning process is the same as in 3D static planning. The 4D dosimetry introduced here uses a different method to delineate PTV (called “IPTV” in this work), which carries all of the 4D information into the planning process. This motion‐weighted IPTV allows generation of motion‐weighted DVH's. Method and Materials: Our new method expands the GTV to a CTV and PTV on each individual respiratory phase and makes a union of all the PTV's to create the IPTV. Different sub‐volumes of the IPTV are assigned different importance in treatment planning. Volumes occupied by PTV throughout the entire respiratory cycle receive higher dosimetric priority than those occupied by PTV only a fraction of the time. Coordinates of the center‐of‐mass of each PTV volume are used to calculate a motion vector for each phase. These motion vectors are used for motion‐weighted dose calculation based upon the 3D dose distribution. Motion‐weighted DVH's are generated from the motion‐weighted dose. The same process can be applied to normal structures. Results: Motion‐weighted dose distribution differs from 3D dose distribution for both target volumes and normal structures. Reducing coverage to volumes of high PTV occupancy worsens the DVH more than reducing coverage to regions of low occupancy. Conclusions: By altering the current ITV to PTV expansion method and weighting the dose with motion, 4D dosimetry can be realized with current planning systems and used to improve treatment plans and generate more realistic DVH's of mobile structures.
36(2009); http://dx.doi.org/10.1118/1.3181102View Description Hide Description
Purpose: Varian RapidArc™ is a new treatment technique that delivers conformal dose distributions to the target in one single gantry rotation. The segments in the treatment arc are allowed to have different monitor unit (MU) weightings requiring variable dose‐rate (VDR) for delivery. An alternative delivery approach for (VDR) RapidArc plans that utilizes constant dose‐rate (CDR) was developed. Method and Materials: For four patient cases (two head‐and‐neck (HN), one brain, one prostate), RapidArc plans were generated using the Eclipse treatment planning system. The resultant VDR treatment plans were converted into CDR plans by re‐distributing the equi‐spaced segments in the VDR arc such that the segments with larger MU weighting occupy a greater angular interval. To minimize perturbation from the optimized dose distribution, the angular deviation of the segments was restricted to 5°. This restriction requires the treatment arc to be broken into multiple sectors such that the local MU fluctuation within each sector is reduced, thereby lowering the angular deviation of the segments during re‐distribution. The converted broken‐sector CDR (b‐CDR) plans were delivered with a single gantry sweep as in the VDR plans but each sector was delivered with different values of CDR. Results and Conclusion: For plans with complex angular MU distribution, the number of broken sectors may increase in the b‐CDR plans in order to maintain the original plan quality. All b‐CDR plans produced similar dose distributions to the VDR plans by dissecting the arc up to 5 sectors. On average, the delivery times of the b‐CDR plans were 1 minute longer than that in the VDR plans. Since the majority of treatment time is spent on the patient setup, these results show that VDR RapidArc plans can be delivered using CDR without significantly compromising the plan quality or treatment efficiency.
Research sponsored by Varian Medical Systems.
36(2009); http://dx.doi.org/10.1118/1.3181103View Description Hide Description
Purpose: To develop a prototype planning system for dynamic Gamma Knife radiosurgery.Method and Materials: Dynamic Gamma Knife is a new concept for implementing Gamma Knife radiosurgery. It uses the spherical high dose volume created by the Gamma Knife unit is as a 3D “paintbrush”, and treatment planning becomes routing the “paintbrush” to “paint” a 3D tumor volume. We have recently finished developing an inverse planning system for dynamic Gamma Knife radiosurgery. Our planning system is a “turn‐key” solution, where the inverse planning problem is solved as a traveling salesman problem combined with constrained least square optimizations. Its input parameters include: (1) contoured anatomy and prescriptions, (2) desired delivery time, (3) APS (advanced patient positioning system) speed, and (4) amount of system memory available. The outputs include: (1) a control point sequence for the APS system together with shot configuration and beam‐on times at each control point, and (2) dose distributions and DVH plots for each structures. Results and Conclusion: We have tested simulated cases that included a spherical tumor, a C‐shaped tumor, and a C‐shaped tumor surrounding a spherical critical structure. The results of these tests showed that: (1) dynamic Gamma Knife radiosurgery is ideally suited for inverse planning, where high quality radiosurgery plans with tight isodose distributions and sharp DVH curves can be obtained in minutes of computation; (2) dynamic radiosurgery plans are more conformal and uniform than current plans and have significantly shorter delivery times; (3) dynamic Gamma Knife radiosurgery can maintain steep dose gradient (around 13% per mm) between the target volume and the surrounding critical structures; (4) with dynamic Gamma Knife radiosurgery, one can obtain a family of plans representing a tradeoff between the delivery time and the dose distributions, thus giving the clinician flexibility in choosing a plan based on the clinical situations.
SU‐EE‐A2‐06: Dosimetric Verification of Ultra Small Fields of Image‐Guided Linac‐Based Stereotactic Radiosurgery36(2009); http://dx.doi.org/10.1118/1.3181104View Description Hide Description
Purpose: To investigate mechanical and dosimetric accuracy of ultra small fields of a cone beam CT (CBCT)‐based stereotactic radiosurgery(SRS) system. Method and Materials: Ultra small fields (7.5 mm, 10 mm, and 12.5 mm cones) of an Elekta linac‐based SRS system were assessed using CBCT‐based setup and filmdosimetry. To examine mechanical accuracy of the linac for SRS, a technique of image analysis was developed using flex maps acquired with a phantom holding a CT‐compatible sphere target at the center. The accuracy of CBCT‐based SRS setup was evaluated with sophisticatedly designed geometrical structures of the phantom in sub‐millimeter precision. For dosimetric verification, another cylindrical acrylic phantom with film inserts was developed. The film phantom has marker structures for setup verification. Radiochromic film was inserted into the phantom and 4 SRS plans of 100° arc beams using three SRS cones were delivered. Dose difference distributions and profiles between computations (Philips Pinnacle3) and the film measurements were qualitatively compared. For quantitative analysis, the computations and the measurements were generated in a resolution of 0.25 mm/pixel and a gamma test with 3%/1mm (dose difference/distance‐to‐agreement (DTA)) criteria was performed. Results: The spherical target in the flex maps (portal images of 0.5 mm/pixel) and image‐guided sphere phantom setup showed the mechanical uncertainty of the SRS system is less than 1 mm. The geometrical marker structures in the film phantom were employed to set up the phantom with sub‐millimeter accuracy using CBCT. The gamma test showed that the dose delivery matched the dose calculation within 3% dose difference or 1mm DTA. Conclusion: The comprehensive quality assurance tools using CBCT‐based SRS setup and radiochromic filmdosimetry allowed us to achieve maximum 1 mm of mechanical uncertainty and 3% of dosimetric uncertainty of ultra small fields in our linac‐based SRS system.
- Moderated Poster ‐ Monte Carlo and General
MO‐EE‐A2‐01: Nuclear Model Evaluation of Uncertainties in Therapeutic Absorbed Dose and Secondary Neutron Production in Proton Radiotherapy Using MCNPX36(2009); http://dx.doi.org/10.1118/1.3182253View Description Hide Description
Purpose: To evaluate differences in nuclear physics models of MCNPX for uncertainties in 1) the dosimetriccharacteristics of the therapeuticdose for SOBP and cross‐field profiles and 2) stray neutron fields produced in a proton therapy unit using the default Bertini model option as the baseline. Methods and Materials: The general purpose Monte Carlo N‐Particle eXtended (MCNPX) code was used to calculate the therapeutic absorbed dose and spectral neutron fluence with three nuclear model options: the Bertini Model, the Cascade Exciton Model (CEM), and the Liège Intra‐Nuclear Cascade Model (INCL4). Each model is included as a physics option in MCNPX. For initial beam energy of 160 MeV, dose calculations were carried out in a water phantom positioned 24 cm downstream of the nozzle exit while in‐air calculations of the neutron spectral fluence were tallied in spheres positioned at isocenter, 100 cm downstream of isocenter and at lateral distances of ± 100 cm off‐axis. Calculations of secondary neutrons produced external to the nozzle housing were evaluated among the models for several parameters including the ambient neutrondose equivalent per therapeutic absorbed dose.Results: Our results indicate that calculations of the therapeutic absorbed dose were in close agreement (⩽1% difference) for all nuclear model options. On the other hand, the differences in the neutron spectral fluence calculations typically varied by a factor of 2 or more among the models. Conclusions: We found that all three inelastic nuclear interaction models provided excellent agreement in the therapeuticdose distributions. However, we observed substantial differences in the neutron spectral fluence calculations that suggest further investigation may be desirable for estimating the out of field radiation exposure uncertainties among the models.
36(2009); http://dx.doi.org/10.1118/1.3182254View Description Hide Description
Purpose: To develop a multileaf collimator(MLC) for proton therapy.Method and Materials: We constructed a model of the IBA proton beam delivery nozzle at our facility with the Geant4 (v9.1) toolkit and, in partnership with Varian Medical Systems, employed it to optimize the shielding design for a multileaf collimator system. Simulation studies were carried out with the highest‐energy, maximally‐modulated spread‐out‐Bragg‐peak settings and for the largest possible field size in both double scattering and uniform scanning modes. This led us iteratively to a solution that maintained proton and neutron leakage doses at an acceptable level. Results: Our final design comprised 50 leaf pairs made from a tungsten‐based alloy. Each leaf is roughly 9 cm high and 11 cm long—enabling 1.5 cm of overtravel—and projects to approximately 0.5 cm at isocenter; one half‐height 450 μm side‐step and two one‐third‐height 300 μm end‐steps are built in. Maximum achievable field sizes are 16×16 cm2 and 18×25 cm2 respectively in double scattering and uniform scanning modes. Conclusion: A MLC system has been designed for use in double scattered and uniform scanned proton therapy without significantly changing the standard IBA nozzle design or the mechanical and electronic systems currently used by Varian photonMLCs. Our simulations predict that the system will provide shielding at least as good as that provided by a conventional brass aperture. A prototype has been manufactured and is presently being installed on the nozzle in one of our gantry rooms. We expect to make measurements of leakage, penumbra and activation in the coming weeks and to validate our Monte Carlo simulations.Disclaimer: This work was supported by the US Army Medical Research and Materiel Command under Contract Agreement No. DAMD17‐W81XWH‐04‐2‐0022. Opinions, interpretations, conclusions and recommendations are those of the author and are not necessarily endorsed by the US Army.
36(2009); http://dx.doi.org/10.1118/1.3182255View Description Hide Description
Purpose: To validate the use of improved source and geometry parameters, derived from Monte Carlo simulation of a Siemens Primus accelerator at maximum field size (40 × 40 cm2), for clinically relevant small electron fields. Method and Materials: Measurements were performed on a Siemens Primus linear accelerator for electron energies 6–21 MeV with open applicators and circular cerrobend inserts ranging from 1 cm to 5 cm in diameter. For the open applicators, relative output factors (ROF), depth penetration and off‐axis profiles were measured with parallel‐plate (Roos) and small volume thimble (CC13) chambers. For the cerrobend insert collimated fields a diode (EFD) was used for high spatial resolution. Off‐axis profiles were measured at 0.5 cm, dmax, in the fall‐off and in the bremsstrahlung tail of each beam. Monte Carlo simulations were performed using the EGSnrc accelerator simulation code BEAMnrc. The source and geometry parameters used were obtained from a previous study following disassembly of the treatment head, for maximum field size and no electron applicator, which resulted in an unprecedented match to measurement. Results: Good agreement was found between measured and calculated off‐axis profiles and percentage depth dose curves, to 2 % / 1 mm or better. The normalized off‐axis profiles in the bremsstrahlung tail matched to 2 – 3 %. Calculated relative output factors were within 1 % of those measured with the Roos chamber. Conclusion: This work has shown that a recent methodology used to extract accurate details on the source and geometry of the treatment head for maximum field size, is applicable for clinically relevant small electron fields. The need for careful measurements and accurate knowledge on the source and geometry of the treatment head as a precursor to accurate fluence and dose calculation is highlighted.
MO‐EE‐A2‐04: Calculation of Water Equivalent Thicknesses of Materials of Arbitrary Densities, Elemental Compositions and Thicknesses in Proton Radiotherapy36(2009); http://dx.doi.org/10.1118/1.3182256View Description Hide Description
Purpose: To develop simple and general analytical formulas to calculate water equivalent thickness (WET) values with an accuracy of 1 mm for materials commonly used in protonradiotherapy.Method and Materials: Based on theoretical range‐energy relationships, three formulas were derived to predict the WET of materials of arbitrary density, composition and thickness Lead,aluminum, polymethylmethacrylate, polystyrene, and lung‐equivalent slabs, which represent materials with different densities, were placed in therapeuticproton fields. Alternative approaches were developed for targets that were ‘radiologically thin’ or ‘thick’. We compared the WET values calculated by the three formulas to those of experimental measurements and iterative numerical calculations. We also compared the accuracy of approximations based on International Atomic Energy Agency (IAEA) report and stopping power ratios (SPR) with numerical calculations. Results: The differences between the values of WET calculated using the analytical formula and the measured data were less than 0.8 mm. Likewise, the WET values calculated using analytical formulas agreed well with those calculated numerically, while IAEA and SPR approximations only work well for thin targets. The comparisons revealed that simple analytical methods could be used to calculate WET values with small, known errors. Conclusion: The results of this study provide evidence that simple analytical formulas can be used to accurately calculate WET of various materials in proton fields of therapeutic energies. Therapeuticproton beams are commonly characterized by their water‐equivalent ranges; therefore, the formulas developed in this study can be used for the convenient and accurate calculation of WET values for common clinical and research applications.
MO‐EE‐A2‐05: Experimental Implementation of the Directed Coulomb Explosion Regime of Laser‐Proton Acceleration36(2009); http://dx.doi.org/10.1118/1.3182257View Description Hide Description
Purpose: To quantify the laser pulse requirements and target parameters required to achieve the Directed Coulomb Explosion (DCE) regime of laser‐target interaction for the acceleration of protons to therapeutic energies. Method and Materials: Particle‐in‐Cell (PIC) simulations of the planned experiments along the funded upgrade path of the HERCULES laser at the University of Michigan have predicted a new regime of attainable laser‐target interactions for proton acceleration. The laser was recently upgraded to 300 TW and a temporal pulse contrast ratio of 10−11, allowing intensities of 2×1022 W/cm2 to be achieved in a near diffraction limited, 1.3 micron, focal spot. The 2 ns long amplified spontaneous emission(ASE) pre‐pulse was suppressed by a factor 10−3 through the implementation of a cross‐polarized wave (XPW) pulse cleaner to prevent pre‐plasma creation on the front surface and preserve the physical integrity of the thin‐film target. Dual plasma mirrors are being characterized to reduce the prepulse at < 30 ps before the main pulse (caused by variations in the index of refraction through the optic path) below a contrast ratio of 10−11. This will allow experiments on thin foil targets (< 100 nm) up to 300TW with no significant pre‐pulse to compromise the target. Results: Preliminary measurements show an additional reduction of the contrast ratio of the ASE pre‐pulse by a factor of 10−3 after the addition of dual plasma mirrors. Additional work is required to optimize the setup and parameters of the plasma mirrors to account for polarization effects and wave front distortions and variable intensity levels. Conclusion: Implementation of dual plasma mirrors is progressing with promising results and will soon allow experimental implementation of the laser pulse characteristics required to test the DCE regime of proton acceleration.
MO‐EE‐A2‐06: Measurements and Comparison of Out‐Of‐Field Organ Doses From Varian Clinac IMRT Plans Using the Atom Phantom36(2009); http://dx.doi.org/10.1118/1.3182258View Description Hide Description
Purpose: To quantify and analyze the unwanted out‐of‐field organdoses during different IMRT prostate treatments and make a systematical comparison of previously reported organdoses.Method and Materials: The measurement involved an ATOM Phantom, thermoluminescent dosimeters(TLD) and a Varian 2100C Clinac. The ATOM phantom was scanned using CT machine to create virtual images for treatment planning. We developed two 6‐MV prostate IMRT plans of 5‐ and 9‐field with the same dose prescriptions, constraints to organs‐at‐risk, and similar in‐field dose distributions. A total of 162 TLDs were placed in the predefined holes to measure the average organdoses of 7 different organs. The results of the two treatment plans were compared to estimate the difference of the out‐of‐field doses.Results: The measurements showed that dose decreases as the distances from the measurement point to the radiation field increases for all treatment plans. The average out‐of‐field organdoses ranged from a maximum of 30.92 cGy to the small intestines to a minimum of 4.40 cGy to the thyroid. The 9‐field IMRT plan resulted in an absorbed dose up to 30% higher than the 5‐field plan which is mainly due to the increased accumulative scatter and lower MU efficiency of the 9‐field IMRT plan. The results were also compared with other studies and the discrepancies were mainly due to the different distances from organs to treatment fields. Conclusion: A method to use TLD and the ATOM physical phantom to measure out‐of‐field organdoses for IMRT has been presented. The measurement results demonstrated that the 9‐field IMRTtreatment delivered higher photondoses than the 5‐field IMRTtreatment to the out‐of‐field organs. Findings from this study involving the prostate treatments express the need to implement the out‐of‐field dose estimation to the treatment planning in order to minimize the secondary cancer and other radiation risk.
- Moderated Poster ‐ Treatment Planning II and Stereotactic
36(2009); http://dx.doi.org/10.1118/1.3182283View Description Hide Description
Purpose: Neural stem cells (NSC) are found in two regions of the adult brain: the subventricular zone (SVZ) of the lateral ventricles and the dentate gyrus (DG) of the hippocampus. These cells are thought to be involved in injury repair and tumor inhibition. This study assesses the feasibility of sparing NSC niches during radiotherapy.Method and Materials: 33 patients treated with IMRT for braintumors were evaluated (10 glioblastoma multiforme (GBM), 10 low grade gliomas, 10 meningiomas, 3 pituitary macroadenomas). IMRTtreatment plans were reconstructed using the Pinnacle treatment planning system (v.8.0m). SVZ and DG were contoured on fused MRI scans. A second IMRT plan using the NSC regions as an avoidance structure was created for each patient. The change in D70 to the contralateral NSC niches and PTV‐1 was recorded. Results: The mean (range) reduction in D70 for the contralateral DG was: 8.3 Gy (0.7–14.3 Gy), 1.8 Gy (−0.15–5.42 Gy), 1.2 Gy (−1.46–7.28 Gy), and 0.5 Gy (−0.4–1.27 Gy) for GBM, low grade gliomas, meningiomas and pituitary adenomas respectively. The mean (range) reduction in D70 for the contralateral SVZ was: 7.3 Gy (2.5–13.4 Gy), 4.2 Gy (−0.5–10.4 Gy), 3.7 Gy (0.43–7.02 Gy), and 5.61 Gy (2–8.34 Gy) for GBM, low grade gliomas, meningiomas and pituitary adenomas respectively. There was no clinically relevant change in the coverage of the PTV‐1. Conclusion: This study demonstrates that it is possible to spare some of the NSC containing regions of the brain using IMRT in patients with braintumors. This is the first study to date to demonstrate feasibility of this technique in numerous patients with multiple tumor histologies.
MO‐FF‐A2‐02: Respiration Induced Heart Motion and Indications of Gated Delivery for Left‐Sided Breast Irradiation36(2009); http://dx.doi.org/10.1118/1.3182284View Description Hide Description
Purpose: To study dosimetric gain of respiratory gating to account for heart motion during left‐sided breast irradiation and to determine indications for gating treatment during treatment planning.Methods and materials: The 4DCT data acquired with free breathing for 13 (out of 68) left‐sided breast cancer patients, who underwent whole breast irradiation with or without regional nodal irradiation, were analyzed retrospectively. Contours of the targets, lung and heart from the planning CT, selected to be the CT at 20% phase, were populated to 0‐ and 50%‐phase CT using deformable registration. The 3D dose distributions were reconstructed in these three phases (0, 20 and 50%). The heart dislocation between the breathing phases was measured in three selected transverse CT slices for the three phases by the changes of DLAD [the distance from left ascending aorta (LAD) to a fixed line drawn on each slice], and maximal heart depth (MDH, the distance of the forefront of the heart to the line). These distances were correlated with the changes of mean heart dose (MHD) and V25.2 for heart between the breathing phases. Results: Significant respiration induced heart displacement was seen, which resulted in substantial variations in dose delivered to the heart. In particular, the heart appeared to move towards to the chest wall during respiration, DLAD changed up to 9 mm, and MDH changed 10.4 mm, 11.0 mm and 10.7mm, respectively, on the three transverse CT slices from superior to inferior. The MHD and V25.2 varied up to 38% and 39%, respectively. These variations were reduced substantially with gating. Conclusion: The respiration induced heart displacement can result in significant variation in heart dose during left‐sided breast irradiation. A large variation in the distances: MDH and/or DLAD, can be used as an indicator to trigger respiratory management, such as gating prior to the treatment delivery.