Neural Spherical Harmonics for structurally coherent continuous representation of diffusion MRI signal

TL;DR

This paper introduces a neural spherical harmonics model for diffusion MRI that captures structural coherence across the brain, improves data reconstruction, noise removal, and enables effective upsampling using a single, hyperparameter-tuned architecture.

Contribution

The paper presents a novel neural spherical harmonics approach (NeSH) for continuous, structurally coherent diffusion MRI representation from a single subject, outperforming existing methods in reconstruction and upsampling.

Findings

More coherent diffusion MRI data representation.

Effective noise removal in gradient images.

Superior upsampling performance.

Abstract

We present a novel way to model diffusion magnetic resonance imaging (dMRI) datasets, that benefits from the structural coherence of the human brain while only using data from a single subject. Current methods model the dMRI signal in individual voxels, disregarding the intervoxel coherence that is present. We use a neural network to parameterize a spherical harmonics series (NeSH) to represent the dMRI signal of a single subject from the Human Connectome Project dataset, continuous in both the angular and spatial domain. The reconstructed dMRI signal using this method shows a more structurally coherent representation of the data. Noise in gradient images is removed and the fiber orientation distribution functions show a smooth change in direction along a fiber tract. We showcase how the reconstruction can be used to calculate mean diffusivity, fractional anisotropy, and total apparent fiber density. These results can be achieved with a single model architecture, tuning only one hyperparameter. In this paper we also demonstrate how upsampling in both the angular and spatial domain yields reconstructions that are on par or better than existing methods.