skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: A nonvoxel-based dose convolution/superposition algorithm optimized for scalable GPU architectures

Abstract

Purpose: Real-time adaptive planning and treatment has been infeasible due in part to its high computational complexity. There have been many recent efforts to utilize graphics processing units (GPUs) to accelerate the computational performance and dose accuracy in radiation therapy. Data structure and memory access patterns are the key GPU factors that determine the computational performance and accuracy. In this paper, the authors present a nonvoxel-based (NVB) approach to maximize computational and memory access efficiency and throughput on the GPU. Methods: The proposed algorithm employs a ray-tracing mechanism to restructure the 3D data sets computed from the CT anatomy into a nonvoxel-based framework. In a process that takes only a few milliseconds of computing time, the algorithm restructured the data sets by ray-tracing through precalculated CT volumes to realign the coordinate system along the convolution direction, as defined by zenithal and azimuthal angles. During the ray-tracing step, the data were resampled according to radial sampling and parallel ray-spacing parameters making the algorithm independent of the original CT resolution. The nonvoxel-based algorithm presented in this paper also demonstrated a trade-off in computational performance and dose accuracy for different coordinate system configurations. In order to find the best balance between the computedmore » speedup and the accuracy, the authors employed an exhaustive parameter search on all sampling parameters that defined the coordinate system configuration: zenithal, azimuthal, and radial sampling of the convolution algorithm, as well as the parallel ray spacing during ray tracing. The angular sampling parameters were varied between 4 and 48 discrete angles, while both radial sampling and parallel ray spacing were varied from 0.5 to 10 mm. The gamma distribution analysis method (γ) was used to compare the dose distributions using 2% and 2 mm dose difference and distance-to-agreement criteria, respectively. Accuracy was investigated using three distinct phantoms with varied geometries and heterogeneities and on a series of 14 segmented lung CT data sets. Performance gains were calculated using three 256 mm cube homogenous water phantoms, with isotropic voxel dimensions of 1, 2, and 4 mm. Results: The nonvoxel-based GPU algorithm was independent of the data size and provided significant computational gains over the CPU algorithm for large CT data sizes. The parameter search analysis also showed that the ray combination of 8 zenithal and 8 azimuthal angles along with 1 mm radial sampling and 2 mm parallel ray spacing maintained dose accuracy with greater than 99% of voxels passing the γ test. Combining the acceleration obtained from GPU parallelization with the sampling optimization, the authors achieved a total performance improvement factor of >175 000 when compared to our voxel-based ground truth CPU benchmark and a factor of 20 compared with a voxel-based GPU dose convolution method. Conclusions: The nonvoxel-based convolution method yielded substantial performance improvements over a generic GPU implementation, while maintaining accuracy as compared to a CPU computed ground truth dose distribution. Such an algorithm can be a key contribution toward developing tools for adaptive radiation therapy systems.« less

Authors:
; ; ; ; ;  [1];  [2]
  1. Department of Radiation Oncology, University of California Los Angeles, 200 Medical Plaza, #B265, Los Angeles, California 90095 (United States)
  2. Department of Radiation Oncology, University of Virginia, 1300 Jefferson Park Avenue, Charlottesville, California 22908 (United States)
Publication Date:
OSTI Identifier:
22409482
Resource Type:
Journal Article
Resource Relation:
Journal Name: Medical Physics; Journal Volume: 41; Journal Issue: 10; Other Information: (c) 2014 American Association of Physicists in Medicine; Country of input: International Atomic Energy Agency (IAEA)
Country of Publication:
United States
Language:
English
Subject:
60 APPLIED LIFE SCIENCES; ACCURACY; ALGORITHMS; ANATOMY; COMPARATIVE EVALUATIONS; COMPUTERIZED SIMULATION; PHANTOMS; PLANNING; RADIATION DOSE DISTRIBUTIONS; RADIATION DOSES; RADIOTHERAPY

Citation Formats

Neylon, J., E-mail: jneylon@mednet.ucla.edu, Sheng, K., Yu, V., Low, D. A., Kupelian, P., Santhanam, A., and Chen, Q. A nonvoxel-based dose convolution/superposition algorithm optimized for scalable GPU architectures. United States: N. p., 2014. Web. doi:10.1118/1.4895822.
Neylon, J., E-mail: jneylon@mednet.ucla.edu, Sheng, K., Yu, V., Low, D. A., Kupelian, P., Santhanam, A., & Chen, Q. A nonvoxel-based dose convolution/superposition algorithm optimized for scalable GPU architectures. United States. doi:10.1118/1.4895822.
Neylon, J., E-mail: jneylon@mednet.ucla.edu, Sheng, K., Yu, V., Low, D. A., Kupelian, P., Santhanam, A., and Chen, Q. Wed . "A nonvoxel-based dose convolution/superposition algorithm optimized for scalable GPU architectures". United States. doi:10.1118/1.4895822.
@article{osti_22409482,
title = {A nonvoxel-based dose convolution/superposition algorithm optimized for scalable GPU architectures},
author = {Neylon, J., E-mail: jneylon@mednet.ucla.edu and Sheng, K. and Yu, V. and Low, D. A. and Kupelian, P. and Santhanam, A. and Chen, Q.},
abstractNote = {Purpose: Real-time adaptive planning and treatment has been infeasible due in part to its high computational complexity. There have been many recent efforts to utilize graphics processing units (GPUs) to accelerate the computational performance and dose accuracy in radiation therapy. Data structure and memory access patterns are the key GPU factors that determine the computational performance and accuracy. In this paper, the authors present a nonvoxel-based (NVB) approach to maximize computational and memory access efficiency and throughput on the GPU. Methods: The proposed algorithm employs a ray-tracing mechanism to restructure the 3D data sets computed from the CT anatomy into a nonvoxel-based framework. In a process that takes only a few milliseconds of computing time, the algorithm restructured the data sets by ray-tracing through precalculated CT volumes to realign the coordinate system along the convolution direction, as defined by zenithal and azimuthal angles. During the ray-tracing step, the data were resampled according to radial sampling and parallel ray-spacing parameters making the algorithm independent of the original CT resolution. The nonvoxel-based algorithm presented in this paper also demonstrated a trade-off in computational performance and dose accuracy for different coordinate system configurations. In order to find the best balance between the computed speedup and the accuracy, the authors employed an exhaustive parameter search on all sampling parameters that defined the coordinate system configuration: zenithal, azimuthal, and radial sampling of the convolution algorithm, as well as the parallel ray spacing during ray tracing. The angular sampling parameters were varied between 4 and 48 discrete angles, while both radial sampling and parallel ray spacing were varied from 0.5 to 10 mm. The gamma distribution analysis method (γ) was used to compare the dose distributions using 2% and 2 mm dose difference and distance-to-agreement criteria, respectively. Accuracy was investigated using three distinct phantoms with varied geometries and heterogeneities and on a series of 14 segmented lung CT data sets. Performance gains were calculated using three 256 mm cube homogenous water phantoms, with isotropic voxel dimensions of 1, 2, and 4 mm. Results: The nonvoxel-based GPU algorithm was independent of the data size and provided significant computational gains over the CPU algorithm for large CT data sizes. The parameter search analysis also showed that the ray combination of 8 zenithal and 8 azimuthal angles along with 1 mm radial sampling and 2 mm parallel ray spacing maintained dose accuracy with greater than 99% of voxels passing the γ test. Combining the acceleration obtained from GPU parallelization with the sampling optimization, the authors achieved a total performance improvement factor of >175 000 when compared to our voxel-based ground truth CPU benchmark and a factor of 20 compared with a voxel-based GPU dose convolution method. Conclusions: The nonvoxel-based convolution method yielded substantial performance improvements over a generic GPU implementation, while maintaining accuracy as compared to a CPU computed ground truth dose distribution. Such an algorithm can be a key contribution toward developing tools for adaptive radiation therapy systems.},
doi = {10.1118/1.4895822},
journal = {Medical Physics},
number = 10,
volume = 41,
place = {United States},
year = {Wed Oct 15 00:00:00 EDT 2014},
month = {Wed Oct 15 00:00:00 EDT 2014}
}
  • The purpose of this study is to evaluate dose prediction errors (DPEs) and optimization convergence errors (OCEs) resulting from use of a superposition/convolution dose calculation algorithm in deliverable intensity-modulated radiation therapy (IMRT) optimization for head-and-neck (HN) patients. Thirteen HN IMRT patient plans were retrospectively reoptimized. The IMRT optimization was performed in three sequential steps: (1) fast optimization in which an initial nondeliverable IMRT solution was achieved and then converted to multileaf collimator (MLC) leaf sequences; (2) mixed deliverable optimization that used a Monte Carlo (MC) algorithm to account for the incident photon fluence modulation by the MLC, whereas a superposition/convolutionmore » (SC) dose calculation algorithm was utilized for the patient dose calculations; and (3) MC deliverable-based optimization in which both fluence and patient dose calculations were performed with a MC algorithm. DPEs of the mixed method were quantified by evaluating the differences between the mixed optimization SC dose result and a MC dose recalculation of the mixed optimization solution. OCEs of the mixed method were quantified by evaluating the differences between the MC recalculation of the mixed optimization solution and the final MC optimization solution. The results were analyzed through dose volume indices derived from the cumulative dose-volume histograms for selected anatomic structures. Statistical equivalence tests were used to determine the significance of the DPEs and the OCEs. Furthermore, a correlation analysis between DPEs and OCEs was performed. The evaluated DPEs were within {+-}2.8% while the OCEs were within 5.5%, indicating that OCEs can be clinically significant even when DPEs are clinically insignificant. The full MC-dose-based optimization reduced normal tissue dose by as much as 8.5% compared with the mixed-method optimization results. The DPEs and the OCEs in the targets had correlation coefficients greater than 0.71, and there was no correlation for the organs at risk. Because full MC-based optimization results in lower normal tissue doses, this method proves advantageous for HN IMRT optimization.« less
  • Purpose: To accelerate dose calculation to interactive rates using highly parallel graphics processing units (GPUs). Methods: The authors have extended their prior work in GPU-accelerated superposition/convolution with a modern dual-source model and have enhanced performance. The primary source algorithm supports both focused leaf ends and asymmetric rounded leaf ends. The extra-focal algorithm uses a discretized, isotropic area source and models multileaf collimator leaf height effects. The spectral and attenuation effects of static beam modifiers were integrated into each source's spectral function. The authors introduce the concepts of arc superposition and delta superposition. Arc superposition utilizes separate angular sampling for themore » total energy released per unit mass (TERMA) and superposition computations to increase accuracy and performance. Delta superposition allows single beamlet changes to be computed efficiently. The authors extended their concept of multi-resolution superposition to include kernel tilting. Multi-resolution superposition approximates solid angle ray-tracing, improving performance and scalability with a minor loss in accuracy. Superposition/convolution was implemented using the inverse cumulative-cumulative kernel and exact radiological path ray-tracing. The accuracy analyses were performed using multiple kernel ray samplings, both with and without kernel tilting and multi-resolution superposition. Results: Source model performance was <9 ms (data dependent) for a high resolution (400{sup 2}) field using an NVIDIA (Santa Clara, CA) GeForce GTX 280. Computation of the physically correct multispectral TERMA attenuation was improved by a material centric approach, which increased performance by over 80%. Superposition performance was improved by {approx}24% to 0.058 and 0.94 s for 64{sup 3} and 128{sup 3} water phantoms; a speed-up of 101-144x over the highly optimized Pinnacle{sup 3} (Philips, Madison, WI) implementation. Pinnacle{sup 3} times were 8.3 and 94 s, respectively, on an AMD (Sunnyvale, CA) Opteron 254 (two cores, 2.8 GHz). Conclusions: The authors have completed a comprehensive, GPU-accelerated dose engine in order to provide a substantial performance gain over CPU based implementations. Real-time dose computation is feasible with the accuracy levels of the superposition/convolution algorithm.« less
  • Background: Commercial treatment planning system Pinnacle3 (Philips, Fitchburg, WI, USA) employs a convolution-superposition algorithm for volumetric-modulated arc radiotherapy (VMAT) optimization and dose calculation. Study of Monte Carlo (MC) dose recalculation of VMAT plans for advanced-stage nasopharyngeal cancers (NPC) is currently limited. Methods: Twenty-nine VMAT prescribed 70Gy, 60Gy, and 54Gy to the planning target volumes (PTVs) were included. These clinical plans achieved with a CS dose engine on Pinnacle3 v9.0 were recalculated by the Monaco TPS v5.0 (Elekta, Maryland Heights, MO, USA) with a XVMC-based MC dose engine. The MC virtual source model was built using the same measurement beam datasetmore » as for the Pinnacle beam model. All MC recalculation were based on absorbed dose to medium in medium (Dm,m). Differences in dose constraint parameters per our institution protocol (Supplementary Table 1) were analyzed. Results: Only differences in maximum dose to left brachial plexus, left temporal lobe and PTV54Gy were found to be statistically insignificant (p> 0.05). Dosimetric differences of other tumor targets and normal organs are found in supplementary Table 1. Generally, doses outside the PTV in the normal organs are lower with MC than with CS. This is also true in the PTV54-70Gy doses but higher dose in the nasal cavity near the bone interfaces is consistently predicted by MC, possibly due to the increased backscattering of short-range scattered photons and the secondary electrons that is not properly modeled by the CS. The straight shoulders of the PTV dose volume histograms (DVH) initially resulted from the CS optimization are merely preserved after MC recalculation. Conclusion: Significant dosimetric differences in VMAT NPC plans were observed between CS and MC calculations. Adjustments of the planning dose constraints to incorporate the physics differences from conventional CS algorithm should be made when VMAT optimization is carried out directly with MC dose engine.« less
  • Purpose: Collapsed-cone convolution/superposition (CCCS) dose calculation is the workhorse for IMRT dose calculation. The authors present a novel algorithm for computing CCCS dose on the modern graphic processing unit (GPU). Methods: The GPU algorithm includes a novel TERMA calculation that has no write-conflicts and has linear computation complexity. The CCCS algorithm uses either tabulated or exponential cumulative-cumulative kernels (CCKs) as reported in literature. The authors have demonstrated that the use of exponential kernels can reduce the computation complexity by order of a dimension and achieve excellent accuracy. Special attentions are paid to the unique architecture of GPU, especially the memorymore » accessing pattern, which increases performance by more than tenfold. Results: As a result, the tabulated kernel implementation in GPU is two to three times faster than other GPU implementations reported in literature. The implementation of CCCS showed significant speedup on GPU over single core CPU. On tabulated CCK, speedups as high as 70 are observed; on exponential CCK, speedups as high as 90 are observed. Conclusions: Overall, the GPU algorithm using exponential CCK is 1000-3000 times faster over a highly optimized single-threaded CPU implementation using tabulated CCK, while the dose differences are within 0.5% and 0.5 mm. This ultrafast CCCS algorithm will allow many time-sensitive applications to use accurate dose calculation.« less
  • Purpose: Recent developments in radiation therapy have been focused on applications of charged particles, especially protons. Over the years several dose calculation methods have been proposed in proton therapy. A common characteristic of all these methods is their extensive computational burden. In the current study we present for the first time, to our best knowledge, a GPU-based PBA for proton dose calculations in Matlab. Methods: In the current study we employed an analytical expression for the protons depth dose distribution. The central-axis term is taken from the broad-beam central-axis depth dose in water modified by an inverse square correction whilemore » the distribution of the off-axis term was considered Gaussian. The serial code was implemented in MATLAB and was launched on a desktop with a quad core Intel Xeon X5550 at 2.67GHz with 8 GB of RAM. For the parallelization on the GPU, the parallel computing toolbox was employed and the code was launched on a GTX 770 with Kepler architecture. The performance comparison was established on the speedup factors. Results: The performance of the GPU code was evaluated for three different energies: low (50 MeV), medium (100 MeV) and high (150 MeV). Four square fields were selected for each energy, and the dose calculations were performed with both the serial and parallel codes for a homogeneous water phantom with size 300×300×300 mm3. The resolution of the PBs was set to 1.0 mm. The maximum speedup of ∼127 was achieved for the highest energy and the largest field size. Conclusion: A GPU-based PB algorithm for proton dose calculations in Matlab was presented. A maximum speedup of ∼127 was achieved. Future directions of the current work include extension of our method for dose calculation in heterogeneous phantoms.« less