C Beltran

A fast GPU-based Monte Carlo simulation of proton transport with detailed modeling of nonelastic interactions.

PURPOSE Very fast Monte Carlo (MC) simulations of proton transport have been implemented recently on graphics processing units (GPUs). However, these MCs usually use simplified models for nonelastic proton-nucleus interactions. Our primary goal is to build a GPU-based proton transport MC with detailed modeling of elastic and nonelastic proton-nucleus collisions. METHODS Using the cuda framework, the authors implemented GPU 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 scattering, multiple scattering, and energy loss straggling, (3) modeling of the intranuclear cascade stage of nonelastic interactions when they occur, (4) simulation of nuclear evaporation, and (5) statistical error estimates on the dose. To validate our MC, the authors performed (1) secondary particle yield calculations in proton collisions with therapeutically relevant nuclei, (2) dose calculations in homogeneous phantoms, (3) recalculations of complex head and neck treatment plans from a commercially available treatment planning system, and compared with (GEANT)4.9.6p2/TOPAS. RESULTS Yields, energy, and angular distributions of secondaries from nonelastic collisions on various nuclei are in good agreement with the (GEANT)4.9.6p2 Bertini and Binary cascade models. The 3D-gamma pass rate at 2%-2 mm for treatment plan simulations is typically 98%. The net computational time on a NVIDIA GTX680 card, including all CPU-GPU data transfers, is ∼ 20 s for 1 × 10(7) proton histories. CONCLUSIONS Our GPU-based MC is the first of its kind to include a detailed nuclear model to handle nonelastic interactions of protons with any nucleus. Dosimetric calculations are in very good agreement with (GEANT)4.9.6p2/TOPAS. Our MC is being integrated into a framework to perform fast routine clinical QA of pencil-beam based treatment plans, and is being used as the dose calculation engine in a clinically applicable MC-based IMPT treatment planning system. The detailed nuclear modeling will allow us to perform very fast linear energy transfer and neutron dose estimates on the GPU.

A GPU-accelerated and Monte Carlo-based intensity modulated proton therapy optimization system.

PURPOSE Conventional spot scanning intensity modulated proton therapy (IMPT) treatment planning systems (TPSs) optimize proton spot weights based on analytical dose calculations. These analytical dose calculations have been shown to have severe limitations in heterogeneous materials. Monte Carlo (MC) methods do not have these limitations; however, MC-based systems have been of limited clinical use due to the large number of beam spots in IMPT and the extremely long calculation time of traditional MC techniques. In this work, the authors present a clinically applicable IMPT TPS that utilizes a very fast MC calculation. METHODS An in-house graphics processing unit (GPU)-based MC dose calculation engine was employed to generate the dose influence map for each proton spot. With the MC generated influence map, a modified least-squares optimization method was used to achieve the desired dose volume histograms (DVHs). The intrinsic CT image resolution was adopted for voxelization in simulation and optimization to preserve spatial resolution. The optimizations were computed on a multi-GPU framework to mitigate the memory limitation issues for the large dose influence maps that resulted from maintaining the intrinsic CT resolution. The effects of tail cutoff and starting condition were studied and minimized in this work. RESULTS For relatively large and complex three-field head and neck cases, i.e., >100,000 spots with a target volume of ∼ 1000 cm(3) and multiple surrounding critical structures, the optimization together with the initial MC dose influence map calculation was done in a clinically viable time frame (less than 30 min) on a GPU cluster consisting of 24 Nvidia GeForce GTX Titan cards. The in-house MC TPS plans were comparable to a commercial TPS plans based on DVH comparisons. CONCLUSIONS A MC-based treatment planning system was developed. The treatment planning can be performed in a clinically viable time frame on a hardware system costing around 45,000 dollars. The fast calculation and optimization make the system easily expandable to robust and multicriteria optimization.

Scanning proton beam therapy reduces normal tissue exposure in pelvic radiotherapy for anal cancer

An inter-comparison planning study between photon beam therapy (IMRT) and scanning proton beam therapy (SPBT) for squamous cell carcinoma of the anus (SCCA) is presented. SPBT plans offer significant reduction (>50%, P=0.008) in doses to small bowel, and bone marrow thereby offering the potential to reduce bowel and hemotoxicities.

Proton therapy dose distribution comparison between Monte Carlo and a treatment planning system for pediatric patients with ependymomaa)

PURPOSE Compare dose distributions for pediatric patients with ependymoma calculated using a Monte Carlo (MC) system and a clinical treatment planning system (TPS). METHODS Plans from ten pediatric patients with ependymoma treated using double scatter proton therapy were exported from the TPS and calculated in our MC system. A field by field comparison of the distal edge (80% and 20%), distal fall off (80% to 20%), field width (50% to 50%), and penumbra (80% to 20%) were examined. In addition, the target dose for the full plan was compared. RESULTS For the 32 fields from the 10 patients, the average differences of distal edge at 80% and 20% on central axis between MC and TPS are -1.9 ± 1.7 mm (p < 0.001) and -0.6 ± 2.3 mm (p = 0.13), respectively. Excluding the fields that ranged out in bone or an air cavity, the 80% difference was -0.9 ± 1.7 mm (p = 0.09). The negative value indicates that MC was on average shallower than TPS. The average difference of the 63 field widths of the 10 patients is -0.7 ± 1.0 mm (p < 0.001), negative indicating on average the MC had a smaller field width. On average, the difference in the penumbra was 2.3 ± 2.1 mm (p < 0.001). The average of the mean clinical target volume dose differences is -1.8% (p = 0.001), negative indicating a lower dose for MC. CONCLUSIONS Overall, the MC system and TPS gave similar results for field width, the 20% distal edge, and the target coverage. For the 80% distal edge and lateral penumbra, there was slight disagreement; however, the difference was less than 2 mm and occurred primarily in highly heterogeneous areas. These differences highlight that the TPS dose calculation cannot be automatically regarded as correct.

A graphics processor-based intranuclear cascade and evaporation simulation

Monte Carlo simulations of the transport of protons in human tissue have been deployed on graphics processing units (GPUs) with impressive results. To provide a more complete treatment of non-elastic nuclear interactions in these simulations, we developed a fast intranuclear cascade-evaporation simulation for the GPU. This can be used to model non-elastic proton collisions on any therapeutically relevant nuclei at incident energies between 20 and 250 MeV. Predictions are in good agreement with Geant4.9.6p2. It takes approximately 2 s to calculate 1 10 6 200 MeV proton- 16 O interactions on a NVIDIA GTX680 GPU. A speed-up factor of 20 relative to one Intel i7-3820 core processor thread was achieved.

Initial clinical experience of postmastectomy intensity modulated proton therapy in patients with breast expanders with metallic ports

PURPOSE The feasibility of proton postmastectomy radiation therapy in patients reconstructed with expanders has not been previously reported, limiting treatment options. We analyzed the dosimetric impact of the metallic port contained within expanders on intensity modulated proton therapy (IMPT) and report our techniques and quality control for treating patients in this setting. METHODS AND MATERIALS Twelve patients with the same expander model underwent 2-field IMPT as part of a prospective registry. All planning dosimetry was checked with an in-house graphic processing unit--based Monte Carlo simulation. Proton ranges through the expander were validated using a sample implant. Dosimetric impact of setup metallic port position uncertainty was evaluated. Pre- and posttreatment photographs were obtained and acute toxicity was graded using the Common Terminology Criteria for Adverse Events, version 4.0. RESULTS Nine patients had bilateral skin-sparing mastectomy with bilateral tissue expander reconstruction, and 3 patients had unilateral skin-sparing mastectomy and reconstruction. The left side was treated in 10 patients and the right side in 2. Target coverage and normal tissue dose uncertainties resulting from the expander were small and clinically acceptable. The maximum physician-assessed acute radiation dermatitis was grade 3 in 1 patient, grade 2 in 5 patients, and grade 1 in 6 patients. CONCLUSIONS Postmastectomy IMPT in breast cancer patients with expanders is feasible and associated with favorable clinical target volume coverage and normal tissue sparing, even when taking into account treatment uncertainties; therefore, these patients should be eligible to participate in clinical trials studying the potential role of proton therapy in breast cancer. We caution, however, that institutions should carry out similar analyses of the physical properties and dosimetric impact of the particular expanders used in their practice before considering IMPT.

Spot-scanned pancreatic stereotactic body proton therapy: A dosimetric feasibility and robustness study

PURPOSE We explored the dosimetric potential of spot-scanned stereotactic body proton therapy (SBPT) for pancreatic cancer. METHODS We compared SBPT to stereotactic body intensity-modulated radiotherapy (SB-IMRT) in 10 patients. We evaluated 3 variables in SBPT planning: (1) 4 and 6 mm spot size; (2) single vs. multi-field optimization (SFO vs. MFO); and (3) optimization target volume (OTV) expansion. Robustness analysis was performed with unidirectional isocenter shifts of ±3 mm in x, y, and z and ±3% stopping power uncertainties. RESULTS SBPT plans had lower V10Gy for the stomach and small and large bowels. Under static robustness, a 5 mm OTV and SFO-6 mm spot size represented the best compromise between target and normal structure. A 4-mm spot-size and 3 mm OTV resulted in significant target underdosing with deformable dose accumulation analysis. CONCLUSIONS This study provides a critical basis for clinical translation of spot size, optimization technique, and OTV expansion for pancreatic SBPT.

Characterization of proton pencil beam scanning and passive beam using a high spatial resolution solid‐state microdosimeter

Purpose: This work aims to characterize a proton pencil beam scanning (PBS) and passive double scattering (DS) systems as well as to measure parameters relevant to the relative biological effectiveness (RBE) of the beam using a silicon on insulator (SOI) microdosimeter with well‐defined 3D sensitive volumes (SV). The dose equivalent downstream and laterally outside of a clinical PBS treatment field was assessed and compared to that of a DS beam. Methods: A novel silicon microdosimeter with well‐defined 3D SVs was used in this study. It was connected to low noise electronics, allowing for detection of lineal energies as low as 0.15 keV/μm. The microdosimeter was placed at various depths in a water phantom along the central axis of the proton beam, and at the distal part of the spread‐out Bragg peak (SOBP) in 0.5 mm increments. The RBE values of the pristine Bragg peak (BP) and SOBP were derived using the measured microdosimetric lineal energy spectra as inputs to the modified microdosimetric kinetic model (MKM). Geant4 simulations were performed in order to verify the calculated depth‐dose distribution from the treatment planning system (TPS) and to compare the simulated dose‐mean lineal energy to the experimental results. Results: For a 131 MeV PBS spot (124.6 mm R90 range in water), the measured dose‐mean lineal energy Symbol increased from 2 keV/μm at the entrance to 8 keV/μm in the BP, with a maximum value of 10 keV/μm at the distal edge. The derived RBE distribution for the PBS beam slowly increased from 0.97 ± 0.14 at the entrance to 1.04 ± 0.09 proximal to the BP, then to 1.1 ± 0.08 in the BP, and steeply rose to 1.57 ± 0.19 at the distal part of the BP. The RBE distribution for the DS SOBP beam was approximately 0.96 ± 0.16 to 1.01 ± 0.16 at shallow depths, and 1.01 ± 0.16 to 1.28 ± 0.17 within the SOBP. The RBE significantly increased from 1.29 ± 0.17 to 1.43 ± 0.18 at the distal edge of the SOBP. Symbol. No Caption available. Conclusions: The SOI microdosimeter with its well‐defined 3D SV has applicability in characterizing proton radiation fields and can measure relevant physical parameters to model the RBE with submillimeter spatial resolution. It has been shown that for a physical dose of 1.82 Gy at the BP, the derived RBE based on the MKM model increased from 1.14 to 1.6 in the BP and its distal part. Good agreement was observed between the experimental and simulation results, confirming the potential application of SOI microdosimeter with 3D SV for quality assurance in proton therapy.

Effects of minimum monitor unit threshold on spot scanning proton plan quality

PURPOSE To investigate the influence of the minimum monitor unit (MU) on the quality of clinical treatment plans for scanned proton therapy. METHODS Delivery system characteristics limit the minimum number of protons that can be delivered per spot, resulting in a min-MU limit. Plan quality can be impacted by the min-MU limit. Two sites were used to investigate the impact of min-MU on treatment plans: pediatric brain tumor at a depth of 5-10 cm; a head and neck tumor at a depth of 1-20 cm. Three-field, intensity modulated spot scanning proton plans were created for each site with the following parameter variations: min-MU limit range of 0.0000-0.0060; and spot spacing range of 2-8 mm. Comparisons were based on target homogeneity and normal tissue sparing. For the pediatric brain, two versions of the treatment planning system were also compared to judge the effects of the min-MU limit based on when it is accounted for in the optimization process (Eclipse v.10 and v.13, Varian Medical Systems, Palo Alto, CA). RESULTS The increase of the min-MU limit with a fixed spot spacing decreases plan quality both in homogeneous target coverage and in the avoidance of critical structures. Both head and neck and pediatric brain plans show a 20% increase in relative dose for the hot spot in the CTV and 10% increase in key critical structures when comparing min-MU limits of 0.0000 and 0.0060 with a fixed spot spacing of 4 mm. The DVHs of CTVs show min-MU limits of 0.0000 and 0.0010 produce similar plan quality and quality decreases as the min-MU limit increases beyond 0.0020. As spot spacing approaches 8 mm, degradation in plan quality is observed when no min-MU limit is imposed. CONCLUSIONS Given a fixed spot spacing of ≤4 mm, plan quality decreases as min-MU increased beyond 0.0020. The effect of min-MU needs to be taken into consideration while planning proton therapy treatments.

Optimizing mini-ridge filter thickness to reduce proton treatment times in a spot-scanning synchrotron system.

PURPOSE Study the contributors to treatment time as a function of Mini-Ridge Filter (MRF) thickness to determine the optimal choice for breath-hold treatment of lung tumors in a synchrotron-based spot-scanning proton machine. METHODS Five different spot-scanning nozzles were simulated in TOPAS: four with MRFs of varying maximal thicknesses (6.15-24.6 mm) and one with no MRF. The MRFs were designed with ridges aligned along orthogonal directions transverse to the beam, with the number of ridges (4-16) increasing with MRF thickness. The material thickness given by these ridges approximately followed a Gaussian distribution. Using these simulations, Monte Carlo data were generated for treatment planning commissioning. For each nozzle, standard and stereotactic (SR) lung phantom treatment plans were created and assessed for delivery time and plan quality. RESULTS Use of a MRF resulted in a reduction of the number of energy layers needed in treatment plans, decreasing the number of synchrotron spills needed and hence the treatment time. For standard plans, the treatment time per field without a MRF was 67.0 ± 0.1 s, whereas three of the four MRF plans had treatment times of less than 20 s per field; considered sufficiently low for a single breath-hold. For SR plans, the shortest treatment time achieved was 57.7 ± 1.9 s per field, compared to 95.5 ± 0.5 s without a MRF. There were diminishing gains in time reduction as the MRF thickness increased. Dose uniformity of the PTV was comparable across all plans; however, when the plans were normalized to have the same coverage, dose conformality decreased with MRF thickness, as measured by the lung V20%. CONCLUSIONS Single breath-hold treatment times for plans with standard fractionation can be achieved through the use of a MRF, making this a viable option for motion mitigation in lung tumors. For stereotactic plans, while a MRF can reduce treatment times, multiple breath-holds would still be necessary due to the limit imposed by the proton extraction time. To balance treatment time and normal tissue dose, the ideal MRF choice was shown to be the thinnest option that is able to achieve the desired breath-hold timing.