Optimising diffusion MRI measurements to maximise microstructural feature sensitivity using Monte-Carlo simulations and estimation theory

Motivation: In diffusion MRI (dMRI), microstructural sensitivity is dependent on the choice of sequence parameters. Goal(s): To develop a parameter optimisation framework that maximises sensitivity to target microstructural features, here applied to a model of spherical restrictions for the pulsed-gradient spin-echo sequence. Approach: We used Monte-Carlo simulations to estimate the signal arising from different microstructure and sequence parameter combinations. Signal estimates were integrated into an optimisation framework to identify a set of dMRI measurements that minimises microstructural estimates' variance . Results: Optimal sequence parameters for test substrates incorporating spherical restrictions converged to a few clusters, suggesting repeated measurements may maximise sensitivity to certain microstructural features. Impact: We present a sequence optimisation framework combining Monte-Carlo simulations with estimation theory. The framework is applicable to a broad range of tissue substrates and dMRI sequences. It identifies a combination of dMRI measurements that maximise sensitivity to target microstructural features.