Scalable and massively parallel Monte Carlo photon transport simulations for heterogeneous computing platforms

Yu, Liang; Nina-Paravecino, Fanny; Kaeli, David; Fang, Qianqian

doi:10.1117/1.jbo.23.1.010504

Cited by 146 publications

(96 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From our tests, MMCL has shown about 2 orders of magnitude speedup compared to our highly optimized single-thread CPU implementation. Although we recognize the 2-to 5-fold speed disadvantage for OpenCL compared to CUDA on NVIDIA hardware, as shown in our previous study, 14 in this work, our decision of prioritizing the development of OpenCL MMC is largely motivated by 1) the open-source ecosystem of OpenCL allowing the developed software to be rapidly and widely disseminated through public software repositories, and 2) the upcoming highperformance GPUs from Intel and AMD being expected to attract more development attention to OpenCL libraries and drivers, likely resulting in boosts in performance.…”

Section: Introductionsupporting

confidence: 44%

“…Moreover, based on our previous observations in voxel-based MC using OpenCL and CUDA on NVIDIA devices, 14 we noticed that CUDA-based MC simulation is about 2-to-5-fold faster than the OpenCL implementation due to driver support differences. As a result, we anticipate that if we further port MMCL to the CUDA programming language, one may achieve further speed improvement, with the resulting software limited to NVIDIA GPUs only.…”

Section: Discussionmentioning

confidence: 68%

“…As discussed in our previous investigation, 14 the high scalability of OpenCL permits simultaneous use of multiple computing devices. This includes both simultaneous use of different generations of GPU architecture and mix between GPUs and CPUs.…”

Section: Simulations On Heterogeneous Computing Platformsmentioning

confidence: 86%

See 2 more Smart Citations

GPU-accelerated mesh-based Monte Carlo photon transport simulations

Fang

Yan

2019

Preprint

Self Cite

View full text Add to dashboard Cite

The mesh-based Monte Carlo (MMC) algorithm is increasingly used as the gold-standard for developing new biophotonics modeling techniques in 3-D complex tissues, including both diffusion-based and various Monte Carlo (MC) based methods. Compared to multi-layered and voxel-based MCs, MMC can utilize tetrahedral meshes to gain improved anatomical accuracy, but also results in higher computational and memory demands. Previous attempts of accelerating MMC using graphics processing units (GPUs) have yielded limited performance improvement and are not publicly available. Here we report a highly efficient MMC -MMCL -using the OpenCL heterogeneous computing framework, and demonstrate a speedup ratio up to 420x compared to state-of-the-art singlethreaded CPU simulations. The MMCL simulator supports almost all advanced features found in our widely disseminated MMC software, such as support for a dozen of complex source forms, wide-field detectors, boundary reflection, photon replay and storing a rich set of detected photon information. Furthermore, this tool supports a wide range of GPUs/CPUs across vendors and is freely available with full source codes and benchmark suites at http://mcx.space/#mmc. c 2019.

show abstract

Section: Introductionsupporting

confidence: 44%

Section: Discussionmentioning

confidence: 68%

See 1 more Smart Citation

GPU-accelerated mesh-based Monte Carlo photon transport simulations

Fang

Yan

2019

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…We used an open source, GPU-accelerated Monte Carlo program (Monte Carlo eXtreme) to compute the detection sensitivity functions for DiFC (25). We modeled the tail as a homogenous 4 mm diameter, 4 cm long cylinder, with voxel size of 250 µm 3 .…”

Section: Monte Carlo Simulationsmentioning

confidence: 99%

In VivoMonitoring of Rare Circulating Tumor Cell and Cluster Dissemination in a Multiple Myeloma Xenograft Model

Patil

Tan

Bartosik

et al. 2019

Preprint

View full text Add to dashboard Cite

We recently developed 'Diffuse in vivo Flow Cytometry' (DiFC), a new pre-clinical research tool for enumerating extremely rare fluorescently-labeled circulating cells directly in vivo. In this paper, we developed a green fluorescent protein (GFP) compatible version of DiFC, and used it to non-invasively monitor the circulating tumor cell (CTC) burden over time in a multiple myeloma disseminated xenograft model. We show that DiFC allowed counting of CTCs at estimated concentrations below 1 cell per mL in peripheral blood with a negligible false alarm rate. DiFC also revealed the presence of CTC clusters in circulation to our knowledge for the first time in this model, and allowed us to calculate their size, kinetics, and frequency of shedding. We anticipate that the unique capabilities of DiFC will have many applications in the study of hematogenous metastasis, and as a powerful complementary methodology to liquid biopsy assays.

show abstract

“…32 This tool can shorten the simulation runtime by several hundreds fold compared to conventional CPU based simulations. 32,33 The remainder of the paper is organized as follows. In the Materials and Methods section, we detail the preprocessing steps to create 4-layer head segmentations from the neurodevelopmental MRI brain atlas library.…”

Section: Introductionmentioning

confidence: 99%

Transcranial Photobiomodulation with Near-Infrared Light from Childhood to Elderliness: Simulation of Dosimetry

Yuan

Cassano

Pias

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Significance: Major depressive disorder (MDD) affects over 40 million US adults in their lifetimes. Transcranial photobiomodulation (t-PBM) has been shown to be effective in treating MDD, but the current treatment dosage does not account for anatomical head and brain changes due to aging. Aim:We study effective t-PBM dosage and its variations across age groups using state-of-the-art Monte Caxrlo (MC) simulations and age-dependent brain atlases ranging between 5 to 85 years of age.Approach: Age-dependent brain models are derived from 18 MRI brain atlases. Two extracranial source positions, F3-F4 and Fp1-Fpz-Fp2 in the EEG 10-20 system, are simulated at five selected wavelengths and energy depositions at two MDD-relevant cortical regions -dorsolateral prefrontal cortex (dlPFC) and ventromedial prefrontal cortex (vmPFC) -are quantified.Results: An overall decrease of energy deposition was found with increasing age. A strong negative correlation between the thickness of extra-cerebral tissues (ECT) and energy deposition, suggesting that increasing ECT thickness over age is primarily responsible for reduced energy delivery. The F3-F4 position appears to be more efficient in reaching dlPFC compared to treating vmPFC via the Fp1-Fpz-Fp2 position.Conclusion: Quantitative simulations revealed age-dependent light delivery across the lifespan of human brains, suggesting the needs for personalized and age-adaptive t-PBM treatment planning.

show abstract

Scalable and massively parallel Monte Carlo photon transport simulations for heterogeneous computing platforms

Cited by 146 publications

References 9 publications

GPU-accelerated mesh-based Monte Carlo photon transport simulations

GPU-accelerated mesh-based Monte Carlo photon transport simulations

In VivoMonitoring of Rare Circulating Tumor Cell and Cluster Dissemination in a Multiple Myeloma Xenograft Model

Transcranial Photobiomodulation with Near-Infrared Light from Childhood to Elderliness: Simulation of Dosimetry

Contact Info

Product

Resources

About