Rotationally-invariant mapping of scalar and orientational metrics of neuronal microstructure with diffusion MRI

TL;DR

This paper introduces a rotationally-invariant framework for estimating neuronal microstructure metrics from diffusion MRI, revealing degeneracies and enabling unbiased whole-brain microstructural mapping.

Contribution

It develops a novel analytical and numerical approach using rotational invariants to accurately estimate microstructural parameters without priors, uncovering parameter degeneracies.

Findings

Identifies multiple parameter solutions fitting the data equally well.

Shows that the correct biophysical solution can be distinguished in some brain regions.

Provides unbiased whole-brain microstructural and orientational maps.

Abstract

We develop a general analytical and numerical framework for estimating intra- and extra-neurite water fractions and diffusion coefficients, as well as neurite orientational dispersion, in each imaging voxel. By employing a set of rotational invariants and their expansion in the powers of diffusion weighting, we analytically uncover the nontrivial topology of the parameter estimation landscape, showing that multiple branches of parameters describe the measurement almost equally well, with only one of them corresponding to the biophysical reality. A comprehensive acquisition shows that the branch choice varies across the brain. Our framework reveals hidden degeneracies in MRI parameter estimation for neuronal tissue, provides microstructural and orientational maps in the whole brain without constraints or priors, and connects modern biophysical modeling with clinical MRI.