Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “PLANNING”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,081 records · Page 60

Automated treatment planning for a dedicated multi-source intra-cranial radiosurgery treatment unit accounting for overlapping structures and dose homogeneity

Purpose: The purpose of this work is to advance the two-step approach for Gamma Knife{sup ®} Perfexion™ (PFX) optimization to account for dose homogeneity and overlap between the planning target volume (PTV) and organs-at-risk (OARs).Methods: In the first step, a geometry-based algorithm is used to quickly select isocentre locations while explicitly accounting for PTV-OARs overlaps. In this approach, the PTV is divided into subvolumes based on the PTV-OARs overlaps and the distance of voxels to the overlaps. Only a few isocentres are selected in the overlap volume, and a higher number of isocentres are carefully selected among voxels that are immediately close to the overlap volume. In the second step, a convex optimization is solved to find the optimal combination of collimator sizes and their radiation duration for each isocentre location.Results: This two-step approach is tested on seven clinical cases (comprising 11 targets) for which the authors assess coverage, OARs dose, and homogeneity index and relate these parameters to the overlap fraction for each case. In terms of coverage, the mean V{sub 99} for the gross target volume (GTV) was 99.8% while the V{sub 95} for the PTV averaged at 94.6%, thus satisfying the clinical objectives of 99% for GTV and 95% for PTV, respectively. The mean relative dose to the brainstem was 87.7% of the prescription dose (with maximum 108%), while on average, 11.3% of the PTV overlapped with the brainstem. The mean beam-on time per fraction per dose was 8.6 min with calibration dose rate of 3.5 Gy/min, and the computational time averaged at 205 min. Compared with previous work involving single-fraction radiosurgery, the resulting plans were more homogeneous with average homogeneity index of 1.18 compared to 1.47.Conclusions: PFX treatment plans with homogeneous dose distribution can be achieved by inverse planning using geometric isocentre selection and mathematical modeling and optimization techniques. The quality of the obtained treatment plans are clinically satisfactory while the homogeneity index is improved compared to conventional PFX plans.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

SU-D-BRD-07: Automatic Patient Data Audit and Plan Quality Check to Support ARIA and Eclipse

Purpose: To ensure patient safety and treatment quality in RT departments that use Varian ARIA and Eclipse, we developed a computer software system and interface functions that allow previously developed electron chart checking (EcCk) methodologies to support these Varian systems. Methods: ARIA and Eclipse store most patient information in its MSSQL database. We studied the contents in the hundreds database tables and identified the data elements used for patient treatment management and treatment planning. Interface functions were developed in both c-sharp and MATLAB to support data access from ARIA and Eclipse servers using SQL queries. These functions and additional data processing functions allowed the existing rules and logics from EcCk to support ARIA and Eclipse. Dose and structure information are important for plan quality check, however they are not stored in the MSSQL database but as files in Varian private formats, and cannot be processed by external programs. We have therefore implemented a service program, which uses the DB Daemon and File Daemon services on ARIA server to automatically and seamlessly retrieve dose and structure data as DICOM files. This service was designed to 1) consistently monitor the data access requests from EcCk programs, 2) translate the requests for ARIA daemon services to obtain dose and structure DICOM files, and 3) monitor the process and return the obtained DICOM files back to EcCk programs for plan quality check purposes. Results: EcCk, which was previously designed to only support MOSAIQ TMS and Pinnacle TPS, can now support Varian ARIA and Eclipse. The new EcCk software has been tested and worked well in physics new start plan check, IMRT plan integrity and plan quality checks. Conclusion: Methods and computer programs have been implemented to allow EcCk to support Varian ARIA and Eclipse systems. This project was supported by a research grant from Varian Medical System.

60 APPLIED LIFE SCIENCES↗

SU-E-J-179: Prediction of Pelvic Nodal Coverage Using Mutual Information Between Cone-Beam and Planning CTs

Purpose: To investigate if pelvic nodal coverage for prostate patients undergoing intensity modulated radiotherapy (IMRT) can be predicted using mutual image information computed between planning and cone-beam CTs (CBCTs). Methods: Four patients with high-risk prostate adenocarcinoma were treated with IMRT on a Varian TrueBeam. Plans were designed such that 95% of the nodal planning target volume (PTV) received the prescription dose of 45 Gy (N=1) or 50.4 Gy (N=3). Weekly CBCTs (N=25) were acquired and the nodal clinical target volumes and organs at risk were contoured by a physician. The percent nodal volume receiving prescription dose was recorded as a ground truth. Using the recorded shifts performed by the radiation therapists at the time of image acquisition, CBCTs were aligned with the planning kVCT. Mutual image information (MI) was calculated between the CBCT and the aligned planning CT within the contour of the nodal PTV. Due to variable CBCT fields-of-view, CBCT images covering less than 90% of the nodal volume were excluded from the analysis, resulting in the removal of eight CBCTs. Results: A correlation coefficient of 0.40 was observed between the MI metric and the percent of the nodal target volume receiving the prescription dose. One patient's CBCTs had clear outliers from the rest of the patients. Upon further investigation, we discovered image artifacts that were present only in that patient's images. When those four images were excluded, the correlation improved to 0.81. Conclusion: This pilot study shows the potential of predicting pelvic nodal dosimetry by computing the mutual image information between planning CTs and patient setup CBCTs. Importantly, this technique does not involve manual or automatic contouring of the CBCT images. Additional patients and more robust exclusion criteria will help validate our findings.

60 APPLIED LIFE SCIENCES↗

SU-E-T-40: Analysis of Composite MVCT Planning Dosimetry with SBRT of Upper Peripheral Lung Cancer

Purpose: Quantitatively evaluate and compare the final adaptive planning doses of upper peripherally located lung SBRT treated with Tomotherapy using 3rd party software tool. Methods: With tumor located in the upper quadrant of lung, a 3rd party software tool was implemented to evaluate the Tomotherapy composite dosimetry created by adaptive fan beam MVCT images described by RTOG 0915 dose criteria (48 Gy / 4 fractions). The composite doses was then summarized with deformable registration in this package with corresponding target and critical structures. The final dosimetry variation, both for target and critical structures, were evaluated in a tabular format and isodose distribution comparisons. Results: Composite SBRT treatment doses were evaluated with adaptive planning. The PTV and several critical structures were mapped/deformed into the package via DICOM from Tomotherapy after the final composite doses were created. Initial plan versus the final composite plan calculated from verification images were compared. The ITV defined by 4D CT and contoured on MVCT images were correlated in patient repositioning. Final composite dose calculated for PTV coverage has shown 0.1–0.17 cGy coverage (0.2–0.4% of prescription dose) variation. Total lung and cord were both less than 0.17 Gy which represented <0.4% difference. All other critical structure were within statistical significance. The adaptive plans justified/included the breathing and motion during the treatment process. Final 95% isotope line coverage from prescription has been met without issues. Conclusions: With lung tumor location in the upper peripheral area, breathing control was not necessary required during SBRT treatment using Tomotherapy technique. Slow fan beam CT provides definitive ITV information and the adaptive composite plan for all fractions were suitable for final dose delivery. The final composite dose calculated with Tomotherapy adaptive tool indicated that the composite dosimetry justified the target location with SBRT delivery, safe with minimum margin of errors.

60 APPLIED LIFE SCIENCES↗

SU-E-T-73: Commissioning of a Treatment Planning System for Proton Spot Scanning

Purpose: A treatment planning system (TPS) was commissioned for clinical use with a fixed beam line proton delivery system. An outline of the data collection, modeling, and verification is provided. Methods: Beam data modeling for proton spot scanning in CMS Xio TPS requires the following measurements: (i) integral depth dose curves (IDDCs); (ii) absolute dose calibration; and (iii) beam spot characteristics. The IDDCs for 18 proton energies were measured using an integrating detector in a single spot field in a water phantom. Absolute scaling of the IDDCs were performed based on ion chamber measurements in mono-energetic 10×10 cm{sup 2} fields in water. Beam spot shapes were measured in air using a flat panel scintillator detector at multiple planes. For beam model verification, more than 45 uniform dose phantom and patient plans were generated. These plans were used to measure range, point dose, and longitudinal and lateral profiles. Tolerances employed for verification are: point dose and longitudinal profiles, ±2%; range, ±1 mm; FWHM for lateral profiles, ±2 mm; and patient plan dose distribution, gamma index of >90% at 3%/3 mm criteria. Results: More than 97% of the point dose measurements out of 115 were within +/-2% with maximum deviation of 3%. 98% of the ranges measured were within 1 mm with maximum deviation of 1.4mm. The normalized depth doses were within 2% at all depths. The maximum error in FWHM of lateral profiles was found to be less than 2mm. For 5 patient plans representing different anatomic sites, a total of 38 planes for 12 beams were analyzed for gamma index with average value of 99% and minimum of 94%. Conclusions: The planning system is successfully commissioned and can be safely deployed for clinical use. Measurements of IDDCs on user beam are highly recommended instead of using standard beam IDDCs.

60 APPLIED LIFE SCIENCES↗

SU-E-T-154: Establishment and Implement of 3D Image Guided Brachytherapy Planning System

Purpose: Cannot observe the dose intuitionally is a limitation of the existing 2D pre-implantation dose planning. Meanwhile, a navigation module is essential to improve the accuracy and efficiency of the implantation. Hence a 3D Image Guided Brachytherapy Planning System conducting dose planning and intra-operative navigation based on 3D multi-organs reconstruction is developed. Methods: Multi-organs including the tumor are reconstructed in one sweep of all the segmented images using the multiorgans reconstruction method. The reconstructed organs group establishs a three-dimensional visualized operative environment. The 3D dose maps of the three-dimentional conformal localized dose planning are calculated with Monte Carlo method while the corresponding isodose lines and isodose surfaces are displayed in a stereo view. The real-time intra-operative navigation is based on an electromagnetic tracking system (ETS) and the fusion between MRI and ultrasound images. Applying Least Square Method, the coordinate registration between 3D models and patient is realized by the ETS which is calibrated by a laser tracker. The system is validated by working on eight patients with prostate cancer. The navigation has passed the precision measurement in the laboratory. Results: The traditional marching cubes (MC) method reconstructs one organ at one time and assembles them together. Compared to MC, presented multi-organs reconstruction method has superiorities in reserving the integrality and connectivity of reconstructed organs. The 3D conformal localized dose planning, realizing the 'exfoliation display' of different isodose surfaces, helps make sure the dose distribution has encompassed the nidus and avoid the injury of healthy tissues. During the navigation, surgeons could observe the coordinate of instruments real-timely employing the ETS. After the calibration, accuracy error of the needle position is less than 2.5mm according to the experiments. Conclusion: The speed and quality of 3D reconstruction, the efficiency in dose planning and accuracy in navigation all can be improved simultaneously.

60 APPLIED LIFE SCIENCES↗

SU-E-T-205: Improving Quality Assurance of HDR Brachytherapy: Verifying Agreement Between Planned and Delivered Dose Distributions Using DICOM RTDose and Advanced Film Dosimetry

Purpose: HDR brachytherapy is undergoing significant development, and quality assurance (QA) checks must keep pace. Current recommendations do not adequately verify delivered against planned dose distributions: This is particularly relevant for new treatment planning system (TPS) calculation algorithms (non TG-43 based), and an era of significant patient-specific plan optimisation. Full system checks are desirable in modern QA recommendations, complementary to device-centric individual tests. We present a QA system incorporating TPS calculation, dose distribution export, HDR unit performance, and dose distribution measurement. Such an approach, more common in external beam radiotherapy, has not previously been reported in the literature for brachytherapy. Methods: Our QA method was tested at 24 UK brachytherapy centres. As a novel approach, we used the TPS DICOM RTDose file export to compare planned dose distribution with that measured using Gafchromic EBT3 films placed around clinical brachytherapy treatment applicators. Gamma analysis was used to compare the dose distributions. Dose difference and distance to agreement were determined at prescription Point A. Accurate film dosimetry was achieved using a glass compression plate at scanning to ensure physically-flat films, simultaneous scanning of known dose films with measurement films, and triple-channel dosimetric analysis. Results: The mean gamma pass rate of RTDose compared to film-measured dose distributions was 98.1% at 3%(local), 2 mm criteria. The mean dose difference, measured to planned, at Point A was -0.5% for plastic treatment applicators and -2.4% for metal applicators, due to shielding not accounted for in TPS. The mean distance to agreement was 0.6 mm. Conclusion: It is recommended to develop brachytherapy QA to include full-system verification of agreement between planned and delivered dose distributions. This is a novel approach for HDR brachytherapy QA. A methodology using advanced film dosimetry and gamma comparison to DICOM RTDose files has been demonstrated as suitable to fulfil this need.

07 ISOTOPE AND RADIATION SOURCES↗

SU-E-T-219: Investigation of IMRT Out-Of-Field Dose Calculation Accuracy for a Commercial Treatment Planning System

Purpose: Inaccuracies in out-of-field calculations could lead to underestimation of dose to organs-at-risk. This study evaluates the dose calculation accuracy of a model-based calculation algorithm at points outside the primary treatment field for an intensity modulated radiation therapy (IMRT) plan using experimental measurements. Methods: The treatment planning system investigated is Varian Eclipse V.10 with Analytical Anisotropic Algorithm (AAA). The IMRT fields investigated are from real patient treatment plans. The doses from a dynamic (DMLC) IMRT brain plan were calculated and compared with measured doses at locations outside the primary treatment fields. Measurements were performed with a MatriXX system (2-D chamber array) placed in solid water. All fields were set vertically incident on the phantom and were 9 cm × 6 cm or smaller. The dose was normalized to the central axis for points up to 15 cm off isocenter. The comparisons were performed at depths of 2, 10, 15, and 20 cm Results: The measurements have shown that AAA calculations underestimate doses at points outside the primary treatment field. The underestimation occurs at 2 cm depth and decreases down to a factor of 2 as depth increases to 20 cm. In low dose (<2% of target dose) regions outside the primary fields the local dose underestimations can be >200% compared to measured doses. Relative to the plan target dose, the measured doses to points outside the field were less than 1% at shallow depths and less than 2% at greater depths. Conclusion: Compared to measurements, the AAA algorithm underestimated the dose at points outside the treatment field with the greatest differences observed at shallow depths. Despite large local dose uncertainties predicted by the treatment planning system, the impact of these uncertainties is expected to be insignificant as doses at these points were less than 1-2% of the prescribed treatment dose.

60 APPLIED LIFE SCIENCES↗

SU-E-T-319: The Effect of Slice Thickness On IMRT Planning

Purpose: The accuracy of volume estimated of a treatment planning system is investigated in this study. In addition, the effect of slice thickness on IMRT planning is also studied. Methods: The accuracy in volume determination was investigated using a water phantom containing various objects with known volumes ranging from 1–100cm{sup 3}. The phantom was scanned with different slice thickness (1–10 mm). The CT data sets were sent to Eclipse TPS for contour delineation and volume calculation. The effect of slice thickness on IMRT planning was studied using a commercial phantom containing four different shaped objects. The phantom was scanned with different slice thickness (1–5 mm). IMRT plans were generated for the different CT datasets to calculate TCP, homogeneity (HI) and conformity indices (CI). Results: The variability of volumes with CT slice thickness was significant especially for small volume structures. The minimum and maximum error in the volume estimation is in the range of −2.3% to 92%. On the other hand, with increasing slice thickness, the PTV mean dose and TCP values decreases. Maximum variation of ∼5% was observed in mean dose and ∼2% in TCP with slice thickness change from 1–5 mm. The relative decrease in target volume receiving 95% of prescribed dose is ∼5% slice thickness change from 1–5 mm. HI increases up to 163% and CI decreases by 4% between 1–5 mm slice thickness change, producing highly inhomogeneous and least conformal plan. Conclusion: Accuracy of volume estimation is dependent on CT slice thickness and the contouring algorithm in a TPS. During TPS commissioning and for all clinical protocols, evaluation of volume should be included to provide the limit of accuracy in DVH calculation. A smaller slice thickness provides superior dosimetry with improved TCP values. Thus, the smallest possible slice thickness should be used for IMRT planning.

07 ISOTOPE AND RADIATION SOURCES↗

SU-E-T-428: Dosimetric Impact of Multileaf Collimator Leaf Width On Single and multiple Isocenter Stereotactic IMRT Treatment Plans for multiple Brain Tumors

Purpose: To assess the impacts that multileaf collimator (MLC) leaf width has on the dose conformity and normal brain tissue doses of single and multiple isocenter stereotactic IMRT (SRT) plans for multiple intracranial tumors. Methods: Fourteen patients with 2–3 targets were studied retrospectively. Patients treated with multiple isocenter treatment plans using 9 to 12 non-coplanar beams per lesion underwent repeat planning using single isocenter and 10 to 12 non-coplanar beams with 2.5mm, 3mm and 5mm MLC leaf widths. Brainlab iPlan treatment planning system for delivery with the 2.5mm MLC served as reference. Identical contour sets and dose-volume constraints were applied. The prescribed dose to each target was 25 Gy to be delivered over 5 fractions with a minimum of 99% dose to cover ≥ 95% of the target volume. Results: The lesions and normal brains ranged in size from 0.11 to 51.67cc (median, 2.75cc) and 1090 to 1641cc (median, 1401cc), respectively. The Paddick conformity index for single and multiple isocenter (2.5mm vs. 3mm and 5mm MLCs) was (0.79±0.08 vs. 0.79±0.07 and 0.77±0.08) and (0.79±0.09 vs. 0.77±0.09 and 0.76±0.08), respectively. The average normal brain volumes receiving 15 Gy for single and multiple isocenter (2.5mm vs. 3mm and 5mm MLCs) were (3.65% vs. 3.95% and 4.09%) and (2.89% vs. 2.91% and 2.92%), respectively. Conclusion: The average dose conformity observed for the different leaf width for single and multiple isocenter plans were similar, throughout. However, the average normal brain volumes receiving 2.5 to 15 Gy were consistently lower for the 2.5mm MLC leaf width, especially for single isocenter plans. The clinical consequences of these integral normal brain tissue doses are still unknown, but employing the use of the 2.5mm MLC option is desirable at sparing normal brain tissue for both single and multiple isocenter cases.

60 APPLIED LIFE SCIENCES↗

SU-E-T-461: Validation of Planning Algorithms in Dynamic Conformal Arc in IPlan Using ArcCHECK and 3DVH

Purpose: To study the effect of limited angular resolution of pencil beam calculation (PBC) on dynamic conformal arc plan (DCAP) in iPlan (BrainLab) using the ArcCHECK sytem and 3DVH software (Sun Nuclear Corporation). Methods: Four DCAPs were generated in iPlan RT Dose 4.5 treatment planning system on the ArcCHECK cylindrical phantom with central planning target volume (PTV). A cylindrical shell structure (SHELL) 2.85cm from phantom surface and 1.5 mm thickness was created to simulate the ArcCHECK diode array. Planned doses were calculated using both Monte Carlo calculation (MCC) and PBC algorithms, and exported to 3DVH software for global and target based comparisons using the 3Dgamma index. Four additional DCAPs were created and calculated on patient CT images and mapped onto the ArcCHECK phantom for measurement using a Varian TrueBeam STx. The measurements were compared against both MC and PB calculation using gamma index analysis. Results: For the ArcCHECK phantom, the dose distribution agreement quantified with 3D-gamma index is better (average-gamma (<γ>)=99.9%vs.79.1% and 96.8%vs45.7%, p=0.0294, 0.0286 for gamma (2mm,2%) and (1mm,1%) criteria respectively using Mann-Whitney U test) in the PTV than in the SHELL. The measurements show better agreement with MCC than the PB (<γ>=100%vs.86.7%, 99.6%vs.72.3%, 85.5%vs.50.8%, p=0.021, 0.026, 0.029 for gamma (3mm,3%), (2mm,2%) and (1mm,1%) criteria using Mann-Whitney U test respectively). The effect due to limited (10 degree) angular resolution of the PBC was observed, and it can be one of the possible reasons for poor agreement between measurement and PB calculation. Conclusion: The PBC of iPlan shows poor peripheral dose calculation accuracy for dynamic conformal arc plans due to limited angular resolution, but it performs well in the area close to target volume without considering heterogeneity. Since the user cannot change the 10 degree angular resolution of PBC, MCC is more appropriate for dynamic conformal arc plans.

60 APPLIED LIFE SCIENCES↗

SU-E-T-630: Commissioning for SRS Planning Systems

Purpose: This study will try to find optimal procedures to collect small fields beam data for commissioning in treatment planning systems (TPS), and to provide a protocol to collect output factors for very small field sizes: 0.5 cm × 0.5 cm to 4.0 cm × 4.0 cm.This will help in determining the correct beam configuration methods in TPS planning intensity modulated radiation therapy (IMRT), and stereotactic radiosurgery SRS using mini multileaf collimation (mMLC). Methods: Data has been collected for a mMLC linear accelerator (linac) Novalis from 0.5 cm × 0.5 cm to 10 cm × 10 cm (its maximum field size). The TPS chosen is BrainLab, Eclipse and Cyberknife. The beam data collected was modeled and imported in the TPS. Verification plans were generated in solid water to confirm the goodness of the data. 3D and IMRT plans on regular CT scans were generated and verified using Mapcheck. All 3D plans with field sizes above 4 cm × 4 cm verified excellent using a distance to agreement of 2 mm and a 2% tolerance. IMRT plans gave an error of -8%. New scans with new detectors have been taken, new field sizes were introduced, and focus has been applied on determining the dosimetric leaf gap. Results: Although this is still a work in progress, this study brings several issues to light: the importance of the correct technique in beam data collection from the correct watertank to the correct detectors. Readings for rectangular fields have to be taken especially for fields which one side is under 4 cm. Conclusion: The use of equivalent square fields will not provide correct readings for the fields with large differences between the length and the width.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

SU-E-T-634: Analysis of Volume Based GYN HDR Brachytherapy Plans for Dose Calculation to Organs At Risk(OAR)

Purpose: We have analyzed the dose volume histogram of 140 CT based HDR brachytherapy plans and evaluated the dose received to OAR ; rectum, bladder and sigmoid colon based on recommendations from ICRU and Image guided brachytherapy working group for cervical cancer . Methods: Our treatment protocol consist of XRT to whole pelvis with 45 Gy at 1.8Gy/fraction followed by 30 Gy at 6 Gy per fraction by HDR brachytherapy in 2 weeks . The CT compatible tandem and ovoid applicators were used and stabilized with radio opaque packing material. The patient was stabilized using special re-locatable implant table and stirrups for reproducibility of the geometry during treatment. The CT scan images were taken at 3mm slice thickness and exported to the treatment planning computer. The OAR structures, bladder, rectum and sigmoid colon were outlined on the images along with the applicators. The prescription dose was targeted to A left and A right as defined in Manchester system and optimized on geometry . The dosimetry was compared on all plans using the parameter Ci.sec.cGy-1 . Using the Dose Volume Histogram (DVH) obtained from the plans the doses to rectum, sigmoid colon and bladder for ICRU defined points and 2cc volume were analyzed and reported. The following criteria were used for limiting the tolerance dose by volume (D2cc) were calculated. The rectum and sigmoid colon doses were limited to <75Gy. The bladder dose was limited to < 90Gy from both XRT and HDR brachytherapy. Results: The average total (XRT+HDRBT) BED values to prescription volume was 120 Gy. Dose 2cc to rectum was 70Gy +/− 17Gy, dose to 2cc bladder was 82+/−32 Gy. The average Ci.sec.cGy-1 calculated for the HDR plans was 6.99 +/− 0.5 Conclusion: The image based treatment planning enabled to evaluati volume based dose to critical structures for clinical interpretation.

60 APPLIED LIFE SCIENCES↗

MO-A-BRD-10: A Fast and Accurate GPU-Based Proton Transport Monte Carlo Simulation for Validating Proton Therapy Treatment Plans

Purpose: To build a GPU-based Monte Carlo (MC) simulation of proton transport with detailed modeling of elastic and non-elastic (NE) protonnucleus interactions, for use in a very fast and cost-effective proton therapy treatment plan verification system. Methods: Using the CUDA framework, we implemented kernels for the following tasks: (1) Simulation of beam spots from our possible scanning nozzle configurations, (2) Proton propagation through CT geometry, taking into account nuclear elastic and multiple scattering, as well as energy straggling, (3) Bertini-style modeling of the intranuclear cascade stage of NE interactions, and (4) Simulation of nuclear evaporation. To validate our MC, we performed: (1) Secondary particle yield calculations in NE collisions with therapeutically-relevant nuclei, (2) Pencil-beam dose calculations in homogeneous phantoms, (3) A large number of treatment plan dose recalculations, and compared with Geant4.9.6p2/TOPAS. A workflow was devised for calculating plans from a commercially available treatment planning system, with scripts for reading DICOM files and generating inputs for our MC. Results: Yields, energy and angular distributions of secondaries from NE collisions on various nuclei are in good agreement with the Geant4.9.6p2 Bertini and Binary cascade models. The 3D-gamma pass rate at 2%–2mm for 70–230 MeV pencil-beam dose distributions in water, soft tissue, bone and Ti phantoms is 100%. The pass rate at 2%–2mm for treatment plan calculations is typically above 98%. The net computational time on a NVIDIA GTX680 card, including all CPU-GPU data transfers, is around 20s for 1×10{sup 7} proton histories. Conclusion: Our GPU-based proton transport MC is the first of its kind to include a detailed nuclear model to handle NE interactions on any nucleus. Dosimetric calculations demonstrate very good agreement with Geant4.9.6p2/TOPAS. Our MC is being integrated into a framework to perform fast routine clinical QA of pencil-beam treatment plans. Hardware for such a system will cost under $5000. This work was funded in part by a grant from Varian Medical Systems, Inc.

61 RADIATION PROTECTION AND DOSIMETRY↗

TH-A-9A-08: Knowledge-Based Quality Control of Clinical Stereotactic Radiosurgery Treatment Plans

Purpose: To develop a quality control tool to reduce stereotactic radiosurgery (SRS) planning variability using models that predict achievable plan quality metrics (QMs) based on individual patient anatomy. Methods: Using a knowledge-based methodology that quantitatively correlates anatomical geometric features to resultant organ-at-risk (OAR) dosimetry, we developed models for predicting achievable OAR dose-volume histograms (DVHs) by training with a cohort of previously treated SRS patients. The DVH-based QMs used in this work are the gradient measure, GM=(3/4pi)^1/3*[V50%^1/3−V100%^1/3], and V10Gy of normal brain. As GM quantifies the total rate of dose fall-off around the planning target volume (PTV), all voxels inside the patient's body contour were treated as OAR for DVH prediction. 35 previously treated SRS plans from our institution were collected; all were planned with non-coplanar volumetric-modulated arc therapy to prescription doses of 12–25 Gy. Of the 35-patient cohort, 15 were used for model training and 20 for model validation. Accuracies of the predictions were quantified by the mean and the standard deviation of the difference between clinical and predicted QMs, δQM=QM-clin−QM-pred. Results: Best agreement between predicted and clinical QMs was obtained when models were built separately for V-PTV<2.5cc and V-PTV>2.5cc. Eight patients trained the V-PTV<2.5cc model and seven patients trained the V-PTV>2.5cc models, respectively. The mean and the standard deviation of δGM were 0.3±0.4mm for the training sets and −0.1±0.6mm for the validation sets, demonstrating highly accurate GM predictions. V10Gy predictions were also highly accurate, with δV10Gy=0.8±0.7cc for the training sets and δV10Gy=0.7±1.4cc for the validation sets. Conclusion: The accuracy of the models in predicting two key SRS quality metrics highlights the potential of this technique for quality control for SRS treatments. Future investigations will seek to determine whether QM variations are due to residual model inaccuracies or true plan quality variations in the clinical sample. Support from Varian Medical Systems.

60 APPLIED LIFE SCIENCES↗

Validation of OSLD and a treatment planning system for surface dose determination in IMRT treatments

Purpose: To evaluate the accuracy of skin dose determination for composite multibeam 3D conformal radiation therapy (3DCRT) and intensity modulated radiation therapy (IMRT) treatments using optically stimulated luminescent dosimeters (OSLDs) and Eclipse treatment planning system. Methods: Surface doses measured by OSLDs in the buildup region for open field 6 MV beams, either perpendicular or oblique to the surface, were evaluated by comparing against dose measured by Markus Parallel Plate (PP) chamber, surface diodes, and calculated by Monte Carlo simulations. The accuracy of percent depth dose (PDD) calculation in the buildup region from the authors’ Eclipse system (Version 10), which was precisely commissioned in the buildup region and was used with 1 mm calculation grid, was also evaluated by comparing to PP chamber measurements and Monte Carlo simulations. Finally, an anthropomorphic pelvic phantom was CT scanned with OSLDs in place at three locations. A planning target volume (PTV) was defined that extended close to the surface. Both an 8 beam 3DCRT and IMRT plan were generated in Eclipse. OSLDs were placed at the CT scanned reference locations to measure the skin doses and were compared to diode measurements and Eclipse calculations. Efforts were made to ensure that the dose comparison was done at the effective measurement points of each detector and corresponding locations in CT images. Results: The depth of the effective measurement point is 0.8 mm for OSLD when used in the buildup region in a 6 MV beam and is 0.7 mm for the authors’ surface diode. OSLDs and Eclipse system both agree well with Monte Carlo and/or Markus PP ion chamber and/or diode in buildup regions in 6 MV beams with normal or oblique incidence and across different field sizes. For the multiple beam 3DCRT plan and IMRT plans, the differences between OSLDs and Eclipse calculations on the surface of the anthropomorphic phantom were within 3% and distance-to-agreement less than 0.3 mm. Conclusions: The authors’ experiment showed that OSLD is an accurate dosimeter for skin dose measurements in complex 3DCRT or IMRT plans. It also showed that an Eclipse system with accurate commissioning of the data in the buildup region and 1 mm calculation grid can calculate surface doses with high accuracy and has a potential to replacein vivo measurements.

60 APPLIED LIFE SCIENCES↗

Multi-GPU implementation of a VMAT treatment plan optimization algorithm

Purpose: Volumetric modulated arc therapy (VMAT) optimization is a computationally challenging problem due to its large data size, high degrees of freedom, and many hardware constraints. High-performance graphics processing units (GPUs) have been used to speed up the computations. However, GPU’s relatively small memory size cannot handle cases with a large dose-deposition coefficient (DDC) matrix in cases of, e.g., those with a large target size, multiple targets, multiple arcs, and/or small beamlet size. The main purpose of this paper is to report an implementation of a column-generation-based VMAT algorithm, previously developed in the authors’ group, on a multi-GPU platform to solve the memory limitation problem. While the column-generation-based VMAT algorithm has been previously developed, the GPU implementation details have not been reported. Hence, another purpose is to present detailed techniques employed for GPU implementation. The authors also would like to utilize this particular problem as an example problem to study the feasibility of using a multi-GPU platform to solve large-scale problems in medical physics. Methods: The column-generation approach generates VMAT apertures sequentially by solving a pricing problem (PP) and a master problem (MP) iteratively. In the authors’ method, the sparse DDC matrix is first stored on a CPU in coordinate list format (COO). On the GPU side, this matrix is split into four submatrices according to beam angles, which are stored on four GPUs in compressed sparse row format. Computation of beamlet price, the first step in PP, is accomplished using multi-GPUs. A fast inter-GPU data transfer scheme is accomplished using peer-to-peer access. The remaining steps of PP and MP problems are implemented on CPU or a single GPU due to their modest problem scale and computational loads. Barzilai and Borwein algorithm with a subspace step scheme is adopted here to solve the MP problem. A head and neck (H and N) cancer case is then used to validate the authors’ method. The authors also compare their multi-GPU implementation with three different single GPU implementation strategies, i.e., truncating DDC matrix (S1), repeatedly transferring DDC matrix between CPU and GPU (S2), and porting computations involving DDC matrix to CPU (S3), in terms of both plan quality and computational efficiency. Two more H and N patient cases and three prostate cases are used to demonstrate the advantages of the authors’ method. Results: The authors’ multi-GPU implementation can finish the optimization process within ∼1 min for the H and N patient case. S1 leads to an inferior plan quality although its total time was 10 s shorter than the multi-GPU implementation due to the reduced matrix size. S2 and S3 yield the same plan quality as the multi-GPU implementation but take ∼4 and ∼6 min, respectively. High computational efficiency was consistently achieved for the other five patient cases tested, with VMAT plans of clinically acceptable quality obtained within 23–46 s. Conversely, to obtain clinically comparable or acceptable plans for all six of these VMAT cases that the authors have tested in this paper, the optimization time needed in a commercial TPS system on CPU was found to be in an order of several minutes. Conclusions: The results demonstrate that the multi-GPU implementation of the authors’ column-generation-based VMAT optimization can handle the large-scale VMAT optimization problem efficiently without sacrificing plan quality. The authors’ study may serve as an example to shed some light on other large-scale medical physics problems that require multi-GPU techniques.

60 APPLIED LIFE SCIENCES↗

SU-E-E-12: Validation of the Implementation of Different CT Scanners in Proton Treatment Planning

Purpose: To quantify the impact of the relationship of CT number and Relative Stopping Power (RSP) among different CT-scanners in the evaluation of dose distribution, for different tumor sites and proton therapy modalities like Pencil Beam and Uniform Scanning. Methods: The same tissue characterization phantom was used to analyze the difference in the X-rays energy spectra of 2 CT scanners of the same vendor. One CT is for planning and the other for treatment verification. The Hounsfield unit (HU) variations and associated dosimetric uncertainty were investigated in uniform scanning and pencil beam treatment plans of different sites. At the same time comparisons of the CT calibration curve were done using the same acquisition protocols. The phantom was imaged on a 16 multi-row CT scanner, with energies of 120 and 140 kVp,currents of 263 and 245 mA and slice thickness of 2 and 3mm respectively. The dosimetric uncertainty of the plans was evaluated in a homogeneus phantom comparing DVHs, gamma index criteria 3% 3mm, and range between them. Results: The variation of HU was within the standard deviation of the average for each tissue substitute. The curves fitted with a bilinear interpolation show a maximum deviation in high density materials like cortical bone and solid water, where the HU values deviated more than 0,6% for the 120kVp protocol. For the others curves the deviation was more than 2% for low densities materials. The gamma index uncertainty criteria passed in all the cases. Conclusion: The HU variations for the different scanners were ≤3%. Lower than the 3,5% uncertainties considered in our treatment planning system. The dosimetric analysis shows that the X-ray spectrum has a small effect on the HU-RSP curve, allowing the use of a single tissue characterization curve in our proton treatment planning system for plans verifications.

60 APPLIED LIFE SCIENCES↗