M Bakhtiari

Using an EPID for patient‐specific VMAT quality assurance

PURPOSE A patient-specific quality assurance (QA) method was developed to verify gantry-specific individual multileaf collimator (MLC) apertures (control points) in volumetric modulated arc therapy (VMAT) plans using an electronic portal imaging device (EPID). METHODS VMAT treatment plans were generated in an Eclipse treatment planning system (TPS). DICOM images from a Varian EPID (aS1000) acquired in continuous acquisition mode were used for pretreatment QA. Each cine image file contains the grayscale image of the MLC aperture related to its specific control point and the corresponding gantry angle information. The TPS MLC file of this RapidArc plan contains the leaf positions for all 177 control points (gantry angles). In-house software was developed that interpolates the measured images based on the gantry angle and overlays them with the MLC pattern for all control points. The 38% isointensity line was used to define the edge of the MLC leaves on the portal images. The software generates graphs and tables that provide analysis for the number of mismatched leaf positions for a chosen distance to agreement at each control point and the frequency in which each particular leaf mismatches for the entire arc. RESULTS Seven patients plans were analyzed using this method. The leaves with the highest mismatched rate were found to be treatment plan dependent. CONCLUSIONS This in-house software can be used to automatically verify the MLC leaf positions for all control points of VMAT plans using cine images acquired by an EPID.

EPID dosimetry for pretreatment quality assurance with two commercial systems

This study compares the EPID dosimetry algorithms of two commercial systems for pretreatment QA, and analyzes dosimetric measurements made with each system alongside the results obtained with a standard diode array. 126 IMRT fields are examined with both EPID dosimetry systems (EPIDose by Sun Nuclear Corporation, Melbourne FL, and Portal Dosimetry by Varian Medical Systems, Palo Alto CA) and the diode array, MapCHECK (also by Sun Nuclear Corporation). Twenty‐six VMAT arcs of varying modulation complexity are examined with the EPIDose and MapCHECK systems. Optimization and commissioning testing of the EPIDose physics model is detailed. Each EPID IMRT QA system is tested for sensitivity to critical TPS beam model errors. Absolute dose gamma evaluation (3%, 3 mm, 10% threshold, global normalization to the maximum measured dose) yields similar results (within 1%–2%) for all three dosimetry modalities, except in the case of off‐axis breast tangents. For these off‐axis fields, the Portal Dosimetry system does not adequately model EPID response, though a previously‐published correction algorithm improves performance. Both MapCHECK and EPIDose are found to yield good results for VMAT QA, though limitations are discussed. Both the Portal Dosimetry and EPIDose algorithms, though distinctly different, yield similar results for the majority of clinical IMRT cases, in close agreement with a standard diode array. Portal dose image prediction may overlook errors in beam modeling beyond the calculation of the actual fluence, while MapCHECK and EPIDose include verification of the dose calculation algorithm, albeit in simplified phantom conditions (and with limited data density in the case of the MapCHECK detector). Unlike the commercial Portal Dosimetry package, the EPIDose algorithm (when sufficiently optimized) allows accurate analysis of EPID response for off‐axis, asymmetric fields, and for orthogonal VMAT QA. Other forms of QA are necessary to supplement the limitations of the Portal Vision Dosimetry system. PACS numbers: 87.53.Bn, 87.53.Jw, 87.53.Kn, 87.55.Qr, 87.56.Fc, 87.57.uq

Optimization of beam angles for intensity modulated radiation therapy treatment planning using genetic algorithm on a distributed computing platform.

Planning intensity modulated radiation therapy (IMRT) treatment involves selection of several angle parameters as well as specification of structures and constraints employed in the optimization process. Including these parameters in the combinatorial search space vastly increases the computational burden, and therefore the parameter selection is normally performed manually by a clinician, based on clinical experience. We have investigated the use of a genetic algorithm (GA) and distributed-computing platform to optimize the gantry angle parameters and provide insight into additional structures, which may be necessary, in the dose optimization process to produce optimal IMRT treatment plans. For an IMRT prostate patient, we produced the first generation of 40 samples, each of five gantry angles, by selecting from a uniform random distribution, subject to certain adjacency and opposition constraints. Dose optimization was performed by distributing the 40-plan workload over several machines running a commercial treatment planning system. A score was assigned to each resulting plan, based on how well it satisfied clinically-relevant constraints. The second generation of 40 samples was produced by combining the highest-scoring samples using techniques of crossover and mutation. The process was repeated until the sixth generation, and the results compared with a clinical (equally-spaced) gantry angle configuration. In the sixth generation, 34 of the 40 GA samples achieved better scores than the clinical plan, with the best plan showing an improvement of 84%. Moreover, the resulting configuration of beam angles tended to cluster toward the patients sides, indicating where the inclusion of additional structures in the dose optimization process may avoid dose hot spots. Additional parameter selection in IMRT leads to a large-scale computational problem. We have demonstrated that the GA combined with a distributed-computing platform can be applied to optimize gantry angle selection within a reasonable amount of time.

Applying graphics processor units to Monte Carlo dose calculation in radiation therapy

We investigate the potential in using of using a graphics processor unit (GPU) for Monte-Carlo (MC)-based radiation dose calculations. The percent depth dose (PDD) of photons in a medium with known absorption and scattering coefficients is computed using a MC simulation running on both a standard CPU and a GPU. We demonstrate that the GPUs capability for massive parallel processing provides a significant acceleration in the MC calculation, and offers a significant advantage for distributed stochastic simulations on a single computer. Harnessing this potential of GPUs will help in the early adoption of MC for routine planning in a clinical environment.

Effect of surface waves on radiotherapy dosimetric measurements in water tanks

The effect of surface waves, generated by moving the scanning arms in water phantoms, on radiation dosimetry is studied. It is shown that in large water tanks, high arm speeds can result in dosimetric errors of up to 5%. The measurements that are started after damping the water waves can result in about a 50% improvement in accuracy of measurements. It is shown that the water surfaces at the start of the measurements have high fluctuations that transform to a steady phase by elapsing time.

SU-FF-J-28: A Guidance System for Optical Patient Alignment During Breast Radiotherapy

Purpose: Breast radiotherapy, particularly IMRT, involves large dose gradients and difficult patient positioning problems. A critical requirement for successful treatment is accurate reproduction of the patients position assumed during CT simulation and planning. We have developed an optical image‐guided technique, which assists in accurately and reproducibly positioning the patient, by displaying her real‐time optical image superimposed on a perspective projection image of her 3D CT data. Methods and Materials: The Single Projection Technique (SPT) accurately determines the 3‐D position and orientation of a camera from a single image acquired of a known model. A calibration jig, composed of ten identifiable reflecting spheres, was constructed and CTimaged to provide this model. To implement our method, a digital photograph of the jig was acquired and the 2D coordinates of the spheres were found. Using this information, 3D CT patient data is projected onto the camerasimaging plane, and is displayed on a monitor, superimposed on the real‐time patient image. This enables the therapist to view both the patients current and desired positions, and guides proper patient positioning. Results: The SPT can determine the position and orientation of the camera to an accuracy of 0.2 cm and 0.3°, respectively. Investigations are ongoing to determine the accuracy and reproducibility of our method in terms of the dose distribution of IMRT plans. Film measurements performed on a breast phantom will allow us to determine the spatial accuracy of isodose curves measured in several planes. Conclusions: We have developed a method to calibrate an optical camerasystem and superimpose a perspective projection of a CTimage on a patients real‐time optical image. Displaying this visual information will assist in accurate setup during breast radiotherapy. Future work will enable us to quantify the setup and dose delivery accuracy of this technique.

SU‐FF‐T‐232: IMRT QA Using a Hybrid Mapcheck / Electronic Portal Dosimetry Environment

Purpose 2D array detector planar dose verification systems play an important role in pre‐treatment quality assurance of IMRT plans. In the present study, we are demonstrating a hybrid IMRT pretreatment QA environment using Varians aS500 & aS1000 electronic portal imaging devices(EPID) in conjunction with Sun Nuclear analysis tools. Materials & Methods 5 patients for prostate [minimally modulated], brain [moderately modulated] and head and neck [heavily modulated], each with IMRT plans generated with Eclipse Treatment Planningsystem were used in this study. The plans were designed for sliding‐window IMRT dose delivery technique on a Varian 6 MV linear accelerator fitted with millennium MLC. An individual verification plan per field was generated from the clinically approved plans. Portal dosimetry treatment plans were also generated in a standard manner assuming 1 CU = 100 cGy. The results were compared with Mapcheck measurements. Results The percentage of detectors passing 3%/3mm criterion as well as the CAX dose measured by the system and the computed plan in absolute dose values were determined for evaluation purposes. For minimal or moderately modulated IMRT plans, both the planar dose verification systems gave comparable results in terms of passing rates [3%/3mm >90%] as well as CAX deviations [<2%]. For heavily modulated head and neck treatment plans, the 3%/3mm passing rates of EPIDs [mean=93.3%] were similar to Mapcheck [mean=85.1%]. For EPID, CAX deviations were not included in this study as the CAX in majority of the cases fell on the penumbra regions of the dual carriages fields. Conclusion: It is possible to use MapCheck analysis software, considering the contemporary gold standard for IMRT QA, with EPID‐acquired portal images of IMRT fields. In case of heavily modulated fields, detector size as well as detector spacing influenced the systems performance with EPID showing better results, as expected.

SU‐DD‐A1‐01: Using an Electronic Portal Imaging Device for Patient‐Specific Volumetric Modulated Arc Therapy Quality Assurance

Purpose: To implement a patient specific Quality Assurance (QA) method to verify gantry specific individual Multi‐Leaf Collimator (MLC) apertures (control points) in Volumetric Modulated Arc Therapy (VMAT) plans using an Electronic Portal Imaging Device (EPID). Method and Materials: RapidArc treatment plans were generated on Eclipse treatment planning system (TPS). DICOM images from a Varian EPID (as 1000) acquired in continuous acquisition mode were used for pretreatment QA. Each DICOM file contains the grey scale image of the MLC aperture related to its specific control point and the corresponding gantry angle. The TPS MLC file of this RapidArc plan contains the leaf positions for all the 177 control points. In‐house software was developed that interpolates the measured images and overlays them with the MLC pattern at all control points. The 50% isodose line was used as the edge of MLCs on the portal images. Results: The software generates graphs and tables that provide analysis for the number of mismatched leaf positions for a chosen distance to agreement (DTA) at each control point and the frequency in which each particular leaf mismatches for the entire arc. Five patients QAs were analyzed using this method. Even for a very complex plan 80% of the active leaves passed the 3 mm DTA criteria. We found that maximum and minimum number of mismatched leaves occurred at a gantry angle of 180° (IEC) and 0° respectively. The leaves with the highest mismatched rate are treatment plan dependent. Conclusion: This in‐house software automatically verifies the MLC leaf positions for all the control points of Rapid Arc plans. The leaf edge positions of the detected segment are compared with the calculated positions specified by the TPS and it was found that number of failures depends on complexity of the treatment plan and gantry angle.

SU‐FF‐T‐173: A New Software for Beam Orientation Optimization in Radiation Therapy Using Genetic Algorithm

Purpose: To investigate the effectiveness of the genetic algorithm (GA) for parallel computation and the optimization of beam orientations and beam weights in radiation therapy.Method and Materials: A unified platform employing existing open‐source software was developed to perform radiationdose calculations, and beam orientation and beam weight optimization on a cluster of computers. This software was tested using several prostate cases. For each case, the set of dose distributions corresponding to each of 72 beams with equally spaced gantry angles were calculated using the VMC++ Monte Carlo algorithm. The GA was then used to optimize the beam angles based on the optimized beam weights/fluence map. The GAs crossover and mutation fractions were varied over several runs. Results: The performance of GA in optimizing the beam orientations strongly depended on the crossover and mutation fractions, with an 80% crossover and 20% mutation rate providing the best results. The scores of plans tended to cluster together in later generations. After several generations, in a 3D conformal planning, the best plan provided a 10% improvement compared to a standard four‐field clinical plan. Conclusion: We have developed a platform for performing dose calculation and beam orientation and beam weight optimization. Evaluations were performed using prostate plans, but may be extended to other cases. The performance of the GA in optimizing the beam orientations depends on the crossover/mutation fraction. For the optimal 80% crossover rate, the clustering of scores for later generations indicates that many beam configurations are good candidates for treatment plans, and they each provide an improvement over a standard plan.

SU‐FF‐T‐16: Evaluation of the Dose Perturbation Caused by Tungsten Shields Within a Fletcher‐Suit Delclos Applicator in Ir‐192 HDR Brachytherapy

Purpose: To evaluate the dose perturbation in bladder and rectal regions caused by tungsten shielding within the ovoids of a Fletch‐Suit Delclos gynecological applicator in Ir‐192 HDR brachytherapy procedures. Method and Materials: A Fletcher‐Suit Delclos (FSD) gynecological applicator, assembled with Tungsten shielded ovoids, was rigidly suspended within a water tank. 28 sheets of gafchromic film were then secured in 6.20 mm increments in regions representative of a patients bladder and rectum. A treatment plan was created (PLATO) to deliver 650 cGy to a user defined pseudo point A. Each film was scanned and analyzed to ascertain its dosimetric distribution. The data was arranged into two three‐dimensional matrices based on the physical dimensions of the films and their locations with respect to a pseudo origin (intersection of the central tandem with the ovoids). Dosimetric data was also extracted from the planning system at points corresponding to specific locations on the gafchromic films. The entire process was then repeated for an FSD applicator without shielding in the ovoids. Results: The measured shielded data was compared to the planning system data (computed) that accounted for the tungsten shielding within the ovoids. It was evident that the measured doses in both the bladder and rectal regions were significantly lower than the computed data predicted. The largest reduction was roughly 24% in the bladder region, and 23% in the rectal region. Similar values were noticed in the unshielded comparison, where the unshielded FSD measurements were compared to computed data not accounting for ovoid shielding. This verifies a consistent discrepancy between the computed predictions and the measured doses.Conclusion: With the current dose tolerances for critical organs being estimated using shielded applicators, an accurate assessment of dose is imperative. This study provides a quantitative estimate for adjusting these accepted standards of bladder and rectal dose.