Multi-GPU configuration of 4D intensity modulated radiation therapy inverse planning using global optimization

We report on the design, implementation and characterization of a multi-graphic processing unit (GPU) computational platform for higher-order optimization in radiotherapy treatment planning. In collaboration with a commercial vendor (Varian Medical Systems, Palo Alto, CA), a research prototype GPU-enabled Eclipse (V13.6) workstation was configured. The hardware consisted of dual 8-core Xeon processors, 256 GB RAM and four NVIDIA Tesla K80 general purpose GPUs. We demonstrate the utility of this platform for large radiotherapy optimization problems through the development and characterization of a parallelized particle swarm optimization (PSO) four dimensional (4D) intensity modulated radiation therapy (IMRT) technique. The PSO engine was coupled to the Eclipse treatment planning system via a vendor-provided scripting interface. Specific challenges addressed in this implementation were (i) data management and (ii) non-uniform memory access (NUMA). For the former, we alternated between parameters over which the computation process was parallelized. For the latter, we reduced the amount of data required to be transferred over the NUMA bridge. The datasets examined in this study were approximately 300 GB in size, including 4D computed tomography images, anatomical structure contours and dose deposition matrices. For evaluation, we created a 4D-IMRT treatment plan for one lung cancer patient and analyzed computation speed while varying several parameters (number of respiratory phases, GPUs, PSO particles, and data matrix sizes). The optimized 4D-IMRT plan enhanced sparing of organs at risk by an average reduction of [Formula: see text] in maximum dose, compared to the clinical optimized IMRT plan, where the internal target volume was used. We validated our computation time analyses in two additional cases. The computation speed in our implementation did not monotonically increase with the number of GPUs. The optimal number of GPUs (five, in our study) is directly related to the hardware specifications. The optimization process took 35 min using 50 PSO particles, 25 iterations and 5 GPUs.

[1]  Carmelo J. A. Bastos Filho,et al.  PSO Efficient Implementation on GPUs Using Low Latency Memory , 2015, IEEE Latin America Transactions.

[2]  Steve B. Jiang,et al.  A GPU OpenCL based cross-platform Monte Carlo dose calculation engine (goMC) , 2015, Physics in medicine and biology.

[3]  Won-Ki Jeong,et al.  Exploiting Massive Parallelism for IndexingMulti-Dimensional Datasets on the GPU , 2015, IEEE Transactions on Parallel and Distributed Systems.

[4]  Xuejun Gu,et al.  Radiotherapy Planning Using an Improved Search Strategy in Particle Swarm Optimization , 2017, IEEE Transactions on Biomedical Engineering.

[5]  László Neumann,et al.  Feature-Preserving Volume Filtering , 2002, VisSym.

[6]  Joao Seco,et al.  Motion management with phase-adapted 4D-optimization , 2010, Physics in medicine and biology.

[7]  Steve Webb,et al.  A margin model to account for respiration-induced tumour motion and its variability , 2008, Physics in medicine and biology.

[8]  Steve B. Jiang,et al.  GPU-based ultrafast IMRT plan optimization , 2009, Physics in medicine and biology.

[9]  Nikos Paragios,et al.  Deformable Medical Image Registration: A Survey , 2013, IEEE Transactions on Medical Imaging.

[10]  Steve B. Jiang,et al.  GPU-based ultra-fast direct aperture optimization for online adaptive radiation therapy , 2010, Physics in medicine and biology.

[11]  Russell C. Eberhart,et al.  A new optimizer using particle swarm theory , 1995, MHS'95. Proceedings of the Sixth International Symposium on Micro Machine and Human Science.

[12]  Steve B. Jiang,et al.  GPU-based high-performance computing for radiation therapy , 2014, Physics in medicine and biology.

[13]  P. Iyengar,et al.  Inverse 4D conformal planning for lung SBRT using particle swarm optimization , 2016, Physics in medicine and biology.

[14]  K. Lam,et al.  Improvement of CT-based treatment-planning models of abdominal targets using static exhale imaging. , 1998, International journal of radiation oncology, biology, physics.

[15]  Xun Jia,et al.  Modeling parameterized geometry in GPU-based Monte Carlo particle transport simulation for radiotherapy , 2016, Physics in medicine and biology.

[16]  Steve B. Jiang,et al.  A new Monte Carlo-based treatment plan optimization approach for intensity modulated radiation therapy , 2015, Physics in medicine and biology.

[17]  Paul J. Keall,et al.  IMRT Treatment Planning on 4D Geometries for the Era of Dynamic MLC Tracking , 2014, Technology in cancer research & treatment.

[18]  Lei Xing,et al.  GPU computing in medical physics: a review. , 2011, Medical physics.

[19]  Mahamod Ismail,et al.  Null Steering of Adaptive Beamforming Using Linear Constraint Minimum Variance Assisted by Particle Swarm Optimization, Dynamic Mutated Artificial Immune System, and Gravitational Search Algorithm , 2014, TheScientificWorldJournal.

[20]  James Kennedy,et al.  Defining a Standard for Particle Swarm Optimization , 2007, 2007 IEEE Swarm Intelligence Symposium.

[21]  Yu Lei,et al.  A self-tuned bat algorithm for optimization in radiation therapy treatment planning , 2014, 15th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD).

[22]  Chia-Feng Juang,et al.  Rule-Based Cooperative Continuous Ant Colony Optimization to Improve the Accuracy of Fuzzy System Design , 2014, IEEE Transactions on Fuzzy Systems.

[23]  Steve B. Jiang,et al.  Residual motion of lung tumors in end-of-inhale respiratory gated radiotherapy based on external surrogates. , 2006, Medical physics.

[24]  N Siauve,et al.  Optimization of the sources in local hyperthermia using a combined finite element-genetic algorithm method , 2004, International journal of hyperthermia : the official journal of European Society for Hyperthermic Oncology, North American Hyperthermia Group.

[25]  Anders Forsgren,et al.  Maximizing the probability of satisfying the clinical goals in radiation therapy treatment planning under setup uncertainty. , 2015, Medical physics.

[26]  Russell H. Taylor,et al.  Towards real-time radiation therapy: GPU accelerated superposition/convolution , 2010, Comput. Methods Programs Biomed..

[27]  Kan Liu,et al.  GPU-Accelerated Parallel Coevolutionary Algorithm for Parameters Identification and Temperature Monitoring in Permanent Magnet Synchronous Machines , 2015, IEEE Transactions on Industrial Informatics.

[28]  Fei Peng,et al.  Multi-GPU implementation of a VMAT treatment plan optimization algorithm. , 2015, Medical physics.

[29]  Glauber Duarte Monteiro,et al.  PSO-GPU: accelerating particle swarm optimization in CUDA-based graphics processing units , 2011, GECCO.