Bayesian Optimization of Sampling Densities in MRI

TL;DR

This paper introduces a Bayesian optimization framework for sampling densities in MRI that significantly accelerates trajectory optimization, reducing computational costs while maintaining competitive performance.

Contribution

It presents a novel Bayesian optimization approach with dimension reduction for MRI sampling, eliminating the need for automatic differentiation and improving speed over traditional methods.

Findings

Optimization speed increased by over 20 times

Performance slightly below state-of-the-art learned trajectories

Method reduces computational complexity and training data requirements

Abstract

Data-driven optimization of sampling patterns in MRI has recently received a significant attention.Following recent observations on the combinatorial number of minimizers in off-the-grid optimization, we propose a framework to globally optimize the sampling densities using Bayesian optimization. Using a dimension reduction technique, we optimize the sampling trajectories more than 20 times faster than conventional off-the-grid methods, with a restricted number of training samples. This method -- among other benefits -- discards the need of automatic differentiation.Its performance is slightly worse than state-of-the-art learned trajectories since it reduces the space of admissible trajectories, but comes with significant computational advantages.Other contributions include: i) a careful evaluation of the distance in probability space to generate trajectories ii) a specific training procedure on families of operators for unrolled reconstruction networks and iii) a gradient projection based scheme for trajectory optimization.