Saved in:
Bibliographic Details
Main Authors: Khole, Prathamesh Pradeep, Petiwala, Zahra Kais, Magesh, Shri Prathaa, Mirafzali, Ehsan, Gupta, Utkarsh, Li, Jing-Rebecca, Ianus, Andrada, Marinescu, Razvan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.01988
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929735157678080
author Khole, Prathamesh Pradeep
Petiwala, Zahra Kais
Magesh, Shri Prathaa
Mirafzali, Ehsan
Gupta, Utkarsh
Li, Jing-Rebecca
Ianus, Andrada
Marinescu, Razvan
author_facet Khole, Prathamesh Pradeep
Petiwala, Zahra Kais
Magesh, Shri Prathaa
Mirafzali, Ehsan
Gupta, Utkarsh
Li, Jing-Rebecca
Ianus, Andrada
Marinescu, Razvan
contents We propose ReMiDi, a novel method for inferring neuronal microstructure as arbitrary 3D meshes using a differentiable diffusion Magnetic Resonance Imaging (dMRI) simulator. We first implemented in PyTorch a differentiable dMRI simulator that simulates the forward diffusion process using a finite-element method on an input 3D microstructure mesh. To achieve significantly faster simulations, we solve the differential equation semi-analytically using a matrix formalism approach. Given a reference dMRI signal $S_{ref}$, we use the differentiable simulator to iteratively update the input mesh such that it matches $S_{ref}$ using gradient-based learning. Since directly optimizing the 3D coordinates of the vertices is challenging, particularly due to ill-posedness of the inverse problem, we instead optimize a lower-dimensional latent space representation of the mesh. The mesh is first encoded into spectral coefficients, which are further encoded into a latent $\textbf{z}$ using an auto-encoder, and are then decoded back into the true mesh. We present an end-to-end differentiable pipeline that simulates signals that can be tuned to match a reference signal by iteratively updating the latent representation $\textbf{z}$. We demonstrate the ability to reconstruct microstructures of arbitrary shapes represented by finite-element meshes, with a focus on axonal geometries found in the brain white matter, including bending, fanning and beading fibers. Our source code is available online.
format Preprint
id arxiv_https___arxiv_org_abs_2502_01988
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ReMiDi: Reconstruction of Microstructure Using a Differentiable Diffusion MRI Simulator
Khole, Prathamesh Pradeep
Petiwala, Zahra Kais
Magesh, Shri Prathaa
Mirafzali, Ehsan
Gupta, Utkarsh
Li, Jing-Rebecca
Ianus, Andrada
Marinescu, Razvan
Image and Video Processing
Graphics
Machine Learning
Medical Physics
We propose ReMiDi, a novel method for inferring neuronal microstructure as arbitrary 3D meshes using a differentiable diffusion Magnetic Resonance Imaging (dMRI) simulator. We first implemented in PyTorch a differentiable dMRI simulator that simulates the forward diffusion process using a finite-element method on an input 3D microstructure mesh. To achieve significantly faster simulations, we solve the differential equation semi-analytically using a matrix formalism approach. Given a reference dMRI signal $S_{ref}$, we use the differentiable simulator to iteratively update the input mesh such that it matches $S_{ref}$ using gradient-based learning. Since directly optimizing the 3D coordinates of the vertices is challenging, particularly due to ill-posedness of the inverse problem, we instead optimize a lower-dimensional latent space representation of the mesh. The mesh is first encoded into spectral coefficients, which are further encoded into a latent $\textbf{z}$ using an auto-encoder, and are then decoded back into the true mesh. We present an end-to-end differentiable pipeline that simulates signals that can be tuned to match a reference signal by iteratively updating the latent representation $\textbf{z}$. We demonstrate the ability to reconstruct microstructures of arbitrary shapes represented by finite-element meshes, with a focus on axonal geometries found in the brain white matter, including bending, fanning and beading fibers. Our source code is available online.
title ReMiDi: Reconstruction of Microstructure Using a Differentiable Diffusion MRI Simulator
topic Image and Video Processing
Graphics
Machine Learning
Medical Physics
url https://arxiv.org/abs/2502.01988