NOZOMI: a GPU-accelerated Monte-Carlo numerical approach for microstructure diffusion imaging

Objective: To develop an open-source computational framework for generating realistic white matter substrates and performing Monte-Carlo diffusion MRI simulations that use periodic boundary conditions to avoid artificial restrictions at substrate boundaries. Approach: A) A novel substrate generation algorithm was developed that ensures microstructure continuity across boundaries by adding reciprocal and replicated axons for boundary-crossing axons during substrate generation. B) A dedicated GPU-accelerated Monte-Carlo diffusion simulation engine was included, with multi-threaded CUDA implementation and simulation domain segmentation for computational efficiency. C) Validation and calibration experiments were done to verify geometric continuity across periodic boundaries, confirm simulation accuracy and precision against analytical solutions across diffusion times, compare against established simulators (Camino, Disimpy, MC/DC) under finite-pulse protocols, and guide simulation configuration. D) Diffusion time dependent axial and radial diffusivities across substrates with varying mean axon diameters and orientation dispersions were evaluated to demonstrate the utility of the framework. Main results: The framework successfully constructed complex biomimetic white matter substrates achieving high packing densities with controlled diameter distributions, orientation dispersions, beading morphology, and myelination. Monte-Carlo simulations reveal that, in both intra- and extra-axonal spaces, axon diameter primarily influences time-dependent radial diffusion, while fiber orientation dispersion primarily impacts axial diffusion. Significance: The developed framework is the first 3D white matter simulation framework to implement periodic boundary conditions for Monte-Carlo simulations without geometric discontinuity, providing approximately four times faster substrate generation than comparable tools, while facilitating controlled investigations of compartment-specific diffusion time dependence of microstructure-diffusion relationships.

Authors

Institutions

Publication Details

Journal
Physics in Medicine and Biology
Published
2026-11-19
DOI
https://doi.org/10.1088/1361-6560/aec4b4
Primary Topic
Advanced Neuroimaging Techniques and Applications
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

NOZOMI: a GPU-accelerated Monte-Carlo numerical approach for microstructure diffusion imaging

Mark D. Does, Kevin D. Harkins, TuanKhai NguyenTran
Physics in Medicine and Biology
Advanced Neuroimaging Techniques and Applications
article

NOZOMI: a GPU-accelerated Monte-Carlo numerical approach for microstructure diffusion imaging

Mark D. Does, Kevin D. Harkins, TuanKhai NguyenTran
article en

Abstract

Objective: To develop an open-source computational framework for generating realistic white matter substrates and performing Monte-Carlo diffusion MRI simulations that use periodic boundary conditions to avoid artificial restrictions at substrate boundaries. Approach: A) A novel substrate generation algorithm was developed that ensures microstructure continuity across boundaries by adding reciprocal and replicated axons for boundary-crossing axons during substrate generation. B) A dedicated GPU-accelerated Monte-Carlo diffusion simulation engine was included, with multi-threaded CUDA implementation and simulation domain segmentation for computational efficiency. C) Validation and calibration experiments were done to verify geometric continuity across periodic boundaries, confirm simulation accuracy and precision against analytical solutions across diffusion times, compare against established simulators (Camino, Disimpy, MC/DC) under finite-pulse protocols, and guide simulation configuration. D) Diffusion time dependent axial and radial diffusivities across substrates with varying mean axon diameters and orientation dispersions were evaluated to demonstrate the utility of the framework. Main results: The framework successfully constructed complex biomimetic white matter substrates achieving high packing densities with controlled diameter distributions, orientation dispersions, beading morphology, and myelination. Monte-Carlo simulations reveal that, in both intra- and extra-axonal spaces, axon diameter primarily influences time-dependent radial diffusion, while fiber orientation dispersion primarily impacts axial diffusion. Significance: The developed framework is the first 3D white matter simulation framework to implement periodic boundary conditions for Monte-Carlo simulations without geometric discontinuity, providing approximately four times faster substrate generation than comparable tools, while facilitating controlled investigations of compartment-specific diffusion time dependence of microstructure-diffusion relationships.

Physics in Medicine and Biology
Vanderbilt University (US)
Advanced Neuroimaging Techniques and Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon