Automated Design of Pulse Sequences for Magnetic Resonance Fingerprinting using Physics-Inspired Optimization

Abstract

Magnetic Resonance Fingerprinting (MRF) is a method to extract quantitative tissue properties such as T 1 and T 2 relaxation rates from arbitrary pulse sequences using conventional magnetic resonance imaging hardware. MRF pulse sequences have thousands of tunable parameters which can be chosen to maximize precision and minimize scan time. Here we perform de novo automated design of MRF pulse sequences by applying physics-inspired optimization heuristics. Our experimental data suggests systematic errors dominate over random errors in MRF scans under clinically-relevant conditions of high undersampling. Thus, in contrast to prior optimization efforts, which focused on statistical error models, we use a cost function based on explicit first-principles simulation of systematic errors arising from Fourier undersampling and phase variation. The resulting pulse sequences display features qualitatively different from previously used MRF pulse sequences and achieve fourfold shorter scan time than prior human-designed sequences of equivalent precision in T 1 and T 2 . Furthermore, the optimization algorithm has discovered the existence of MRF pulse sequences with intrinsic robustness against shading artifacts due to phase variation. SignificanceMagnetic resonance is a widely used non-invasive medical imaging technology. Most clinical magnetic resonance imaging scans generate qualitative or 'weighted' images. A recent new technology, Magnetic Resonance Fingerprinting [1], simultaneously extracts quantitative voxel-by-voxel measurements of multiple intrinsic tissue properties such as T 1 and T 2 relaxation rates in a single scan, * The work of H. G. K. was performed before joining Amazon Web Services.