Dual-grid mesh-based Monte Carlo algorithm for efficient photon transport simulations in complex three-dimensional media

Abstract

Fang, "Dual-grid mesh-based Monte Carlo algorithm for efficient photon transport simulations in complex three-dimensional media," J.Abstract. The mesh-based Monte Carlo (MMC) method is an efficient algorithm to model light propagation inside tissues with complex boundaries, but choosing appropriate mesh density can be challenging. A fine mesh improves the spatial resolution of the output but requires more computation. We propose an improved MMCdual-grid mesh-based Monte Carlo (DMMC)-to accelerate photon simulations using a coarsely tessellated tetrahedral mesh for ray-tracing computation and an independent voxelated grid for output data storage. The decoupling between ray-tracing and data storage grids allows us to simultaneously achieve faster simulations and improved output spatial accuracy. Furthermore, we developed an optimized ray-tracing technique to eliminate unnecessary ray-tetrahedron intersection tests in optically thick mesh elements. We validate the proposed algorithms using a complex heterogeneous domain and compare the solutions with those from MMC and voxel-based Monte Carlo. We found that DMMC with an unrefined constrained Delaunay tessellation of the boundary nodes yielded the highest speedup, ranging from 1.3× to 2.9× for various scattering settings, with nearly no loss in accuracy. In addition, the optimized ray-tracing technique offers excellent acceleration in high-scattering media, reducing the raytetrahedron test count by over 100-fold. Our DMMC software can be downloaded at http://mcx.space/mmc. © The Authors. Published by SPIE under a Creative Commons Attribution 4.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

GPU-accelerated mesh-based Monte Carlo photon transport simulations

“…In Fig. 2(b), The previously observed fluence differences inside the spherical shell between mesh-based simulations and voxel-based MC simulations (MCX-CL) 10 are also reproduced in MMCL outputs, emphasizing the importance of using mesh models when simulating domains with curved boundaries and highcontrast heterogeneities.…”

“…In the B2 benchmark, the optical properties and dimensions of each layer are described previously. 10 In both cases, a pencil beam source injects photons at [30, 30, 0] mm, with the initial direction pointing along the +z-axis. Boundary reflection is considered in B2 but not in B1.…”

“…Recently, we reported a significantly improved MMC algorithm, the dual-grid MMC or DMMC, 10 combining the strengths of both voxel-and mesh-based simulations. It reduces the ray-tracing overhead significantly by utilizing a coarse tessellation of the surface nodes without losing anatomical accuracy while a dense voxelated storage grid provides higher spatial resolution and lower discretization error.…”

“…Here, we report a highly accelerated MMC imple-mentation, MMCL, developed using the Open Computing Language (OpenCL) framework. MMCL supports almost all advanced simulation features seen in our open-source CPU MMC, such as the support of wide-field complex sources and detectors, 9 photon replay, 17 dual-grid simulations 10 and storage capabilities for rich sets of detected photon data. Thanks to the excellent portability of OpenCL, the MMCL is capable of running on a wide range of commodity GPUs.…”

Improving model-based fNIRS analysis using mesh-based anatomical and light-transport models

“…The three brain anatomical models that we are evaluating include 1) mesh-based brain models, 6,16 2) voxelbased brain segmentations, 45 and 3) simple layered-slab brain models often found in literature. 3,46 We apply our widely adopted and cross-validated MC light transport simulators -Mesh-based MC (MMC) 16,17 for mesh-and layered-slab brain simulations, and Monte Carlo eXtreme (MCX) 31,47 for voxel-based brain simulations. Furthermore, a number of publications applied the diffusion approximation (DA) to fNIRS modeling by utilizing an approximated scattering coefficient for CSF, 5,48 as it has very low scattering.…”