Hardware acceleration of a Monte Carlo simulation for photodynamic therapy [corrected] treatment planning

GPU-accelerated Monte Carlo simulation of MV-CBCT

Monte Carlo simulation (MCS) is one of the most accurate computation methods for dose calculation and image formation in radiation therapy. However, the high computational complexity and long execution time of MCS limits its broad use. In this paper, we present a novel strategy to accelerate MCS using a graphic processing unit (GPU), and we demonstrate the application in mega-voltage (MV) cone-beam computed tomography (CBCT) simulation. A new framework that generates a series of MV projections from a single simulation run is designed specifically for MV-CBCT acquisition. A Geant4-based GPU code for photon simulation is incorporated into the framework for the simulation of photon transport through a phantom volume. The FastEPID method, which accelerates the simulation of MV images, is modified and integrated into the framework. The proposed GPU-based simulation strategy was tested for its accuracy and efficiency in a Catphan 604 phantom and an anthropomorphic pelvis phantom with beam energies at 2.5 MV, 6 MV, and 6 MV FFF. In all cases, the proposed GPU-based simulation demonstrated great simulation accuracy and excellent agreement with measurement and CPU-based simulation in terms of reconstructed image qualities. The MV-CBCT simulation was accelerated by factors of roughly 900-2300 using an NVIDIA Tesla V100 GPU card against a 2.5 GHz AMD Opteron™ Processor 6380.

Efficient inversion strategies for estimating optical properties with Monte Carlo radiative transport models

SIGNIFICANCE

Indirect imaging problems in biomedical optics generally require repeated evaluation of forward models of radiative transport, for which Monte Carlo is accurate yet computationally costly. We develop an approach to reduce this bottleneck, which has significant implications for quantitative tomographic imaging in a variety of medical and industrial applications.

AIM

Our aim is to enable computationally efficient image reconstruction in (hybrid) diffuse optical modalities using stochastic forward models.

APPROACH

Using Monte Carlo, we compute a fully stochastic gradient of an objective function for a given imaging problem. Leveraging techniques from the machine learning community, we then adaptively control the accuracy of this gradient throughout the iterative inversion scheme to substantially reduce computational resources at each step.

RESULTS

For example problems of quantitative photoacoustic tomography and ultrasound-modulated optical tomography, we demonstrate that solutions are attainable using a total computational expense that is comparable to (or less than) that which is required for a single high-accuracy forward run of the same Monte Carlo model.

CONCLUSIONS

This approach demonstrates significant computational savings when approaching the full nonlinear inverse problem of optical property estimation using stochastic methods.

FullMonteCUDA: a fast, flexible, and accurate GPU-accelerated Monte Carlo simulator for light propagation in turbid media

Optimizing light delivery for photodynamic therapy, quantifying tissue optical properties or reconstructing 3D distributions of sources in bioluminescence imaging and absorbers in diffuse optical imaging all involve solving an inverse problem. This can require thousands of forward light propagation simulations to determine the parameters to optimize treatment, image tissue or quantify tissue optical properties, which is time-consuming and computationally expensive. Addressing this problem requires a light propagation simulator that produces results quickly given modelling parameters. In previous work, we developed FullMonteSW: currently the fastest, tetrahedral-mesh, Monte Carlo light propagation simulator written in software. Additional software optimizations showed diminishing performance improvements, so we investigated hardware acceleration methods. This work focuses on FullMonteCUDA: a GPU-accelerated version of FullMonteSW which targets NVIDIA GPUs. FullMonteCUDA has been validated across several benchmark models and, through various GPU-specific optimizations, achieves a 288-936x speedup over the single-threaded, non-vectorized version of FullMonteSW and a 4-13x speedup over the highly optimized, hand-vectorized and multi-threaded version. The increase in performance allows inverse problems to be solved more efficiently and effectively.

Monte Carlo simulation of photon migration in a cloud computing environment with MapReduce

Monte Carlo simulation is considered the most reliable method for modeling photon migration in heterogeneous media. However, its widespread use is hindered by the high computational cost. The purpose of this work is to report on our implementation of a simple MapReduce method for performing fault-tolerant Monte Carlo computations in a massively-parallel cloud computing environment. We ported the MC321 Monte Carlo package to Hadoop, an open-source MapReduce framework. In this implementation, Map tasks compute photon histories in parallel while a Reduce task scores photon absorption. The distributed implementation was evaluated on a commercial compute cloud. The simulation time was found to be linearly dependent on the number of photons and inversely proportional to the number of nodes. For a cluster size of 240 nodes, the simulation of 100 billion photon histories took 22 min, a 1258 × speed-up compared to the single-threaded Monte Carlo program. The overall computational throughput was 85,178 photon histories per node per second, with a latency of 100 s. The distributed simulation produced the same output as the original implementation and was resilient to hardware failure: the correctness of the simulation was unaffected by the shutdown of 50% of the nodes.

Dedicated hardware processor and corresponding system-on-chip design for real-time laser speckle imaging

Laser speckle imaging (LSI) is a noninvasive and full-field optical imaging technique which produces two-dimensional blood flow maps of tissues from the raw laser speckle images captured by a CCD camera without scanning. We present a hardware-friendly algorithm for the real-time processing of laser speckle imaging. The algorithm is developed and optimized specifically for LSI processing in the field programmable gate array (FPGA). Based on this algorithm, we designed a dedicated hardware processor for real-time LSI in FPGA. The pipeline processing scheme and parallel computing architecture are introduced into the design of this LSI hardware processor. When the LSI hardware processor is implemented in the FPGA running at the maximum frequency of 130 MHz, up to 85 raw images with the resolution of 640×480 pixels can be processed per second. Meanwhile, we also present a system on chip (SOC) solution for LSI processing by integrating the CCD controller, memory controller, LSI hardware processor, and LCD display controller into a single FPGA chip. This SOC solution also can be used to produce an application specific integrated circuit for LSI processing.

Next-generation acceleration and code optimization for light transport in turbid media using GPUs

A highly optimized Monte Carlo (MC) code package for simulating light transport is developed on the latest graphics processing unit (GPU) built for general-purpose computing from NVIDIA - the Fermi GPU. In biomedical optics, the MC method is the gold standard approach for simulating light transport in biological tissue, both due to its accuracy and its flexibility in modelling realistic, heterogeneous tissue geometry in 3-D. However, the widespread use of MC simulations in inverse problems, such as treatment planning for PDT, is limited by their long computation time. Despite its parallel nature, optimizing MC code on the GPU has been shown to be a challenge, particularly when the sharing of simulation result matrices among many parallel threads demands the frequent use of atomic instructions to access the slow GPU global memory. This paper proposes an optimization scheme that utilizes the fast shared memory to resolve the performance bottleneck caused by atomic access, and discusses numerous other optimization techniques needed to harness the full potential of the GPU. Using these techniques, a widely accepted MC code package in biophotonics, called MCML, was successfully accelerated on a Fermi GPU by approximately 600x compared to a state-of-the-art Intel Core i7 CPU. A skin model consisting of 7 layers was used as the standard simulation geometry. To demonstrate the possibility of GPU cluster computing, the same GPU code was executed on four GPUs, showing a linear improvement in performance with an increasing number of GPUs. The GPU-based MCML code package, named GPU-MCML, is compatible with a wide range of graphics cards and is released as an open-source software in two versions: an optimized version tuned for high performance and a simplified version for beginners (http://code.google.com/p/gpumcml).

Photodynamic therapy: superficial and interstitial illumination

Photodynamic therapy (PDT) is reviewed using the treatment of skin tumors as an example of superficial lesions and prostate cancer as an example of deep-lying lesions requiring interstitial intervention. These two applications are among the most commonly studied in oncological PDT, and illustrate well the different challenges facing the two modalities of PDT-superficial and interstitial. They thus serve as good examples to illustrate the entire field of PDT in oncology. PDT is discussed based on the Lund University group's over 20 yr of experience in the field. In particular, the interplay between optical diagnostics and dosimetry and the delivery of the therapeutic light dose are highlighted. An interactive multiple-fiber interstitial procedure to deliver the required therapeutic dose based on the assessment of light fluence rate and sensitizer concentration and oxygen level throughout the tumor is presented.

Solution of the direct problem in turbid media with inclusions using Monte Carlo simulations implemented in graphics processing units: new criterion for processing transmittance data

The study of light propagation in diffusive media requires solving the radiative transfer equation, or eventually, the diffusion approximation. Except for some cases involving simple geometries, the problem with immersed inclusions has not been solved. Also, Monte Carlo (MC) calculations have become a gold standard for simulating photon migration in turbid media, although they have the drawback large processing times. The purpose of this work is two-fold: first, we introduce a new processing criterion to retrieve information about the location and shape of absorbing inclusions based on normalization to the background intensity, when no inhomogeneities are present. Second, we demonstrate the feasibility of including inhomogeneities in MC simulations implemented in graphics processing units, achieving large acceleration factors ( approximately 10(3)), thus providing an important tool for iteratively solving the forward problem to retrieve the optical properties of the inclusion. Results using a cw source are compared with MC outcomes showing very good agreement.

A tetrahedron-based inhomogeneous Monte Carlo optical simulator

Optical imaging has been widely applied in preclinical and clinical applications. Fifteen years ago, an efficient Monte Carlo program 'MCML' was developed for use with multi-layered turbid media and has gained popularity in the field of biophotonics. Currently, there is an increasingly pressing need for simulating tools more powerful than MCML in order to study light propagation phenomena in complex inhomogeneous objects, such as the mouse. Here we report a tetrahedron-based inhomogeneous Monte Carlo optical simulator (TIM-OS) to address this issue. By modeling an object as a tetrahedron-based inhomogeneous finite-element mesh, TIM-OS can determine the photon-triangle interaction recursively and rapidly. In numerical simulation, we have demonstrated the correctness and efficiency of TIM-OS.

Monte Carlo simulation of photon migration in 3D turbid media accelerated by graphics processing units

We report a parallel Monte Carlo algorithm accelerated by graphics processing units (GPU) for modeling time-resolved photon migration in arbitrary 3D turbid media. By taking advantage of the massively parallel threads and low-memory latency, this algorithm allows many photons to be simulated simultaneously in a GPU. To further improve the computational efficiency, we explored two parallel random number generators (RNG), including a floating-point-only RNG based on a chaotic lattice. An efficient scheme for boundary reflection was implemented, along with the functions for time-resolved imaging. For a homogeneous semi-infinite medium, good agreement was observed between the simulation output and the analytical solution from the diffusion theory. The code was implemented with CUDA programming language, and benchmarked under various parameters, such as thread number, selection of RNG and memory access pattern. With a low-cost graphics card, this algorithm has demonstrated an acceleration ratio above 300 when using 1792 parallel threads over conventional CPU computation. The acceleration ratio drops to 75 when using atomic operations. These results render the GPU-based Monte Carlo simulation a practical solution for data analysis in a wide range of diffuse optical imaging applications, such as human brain or small-animal imaging.