Pattern Search Pulse Sequence Optimization for Cardiac MR Fingerprinting

Motivation: Current cardiac MRF sequences have been designed empirically, and thus do not exploit the full flexibility in sequence design of the MRF framework. Goal(s): The goal of this project is to develop a sequence optimization method for cardiac MRF, which can be applied to shorten the breathhold and diastolic window. Approach: An optimization method was implemented that simulates MRI signal generation, spiral undersampling, and pattern recognition during each iteration, using a pattern search algorithm to update multiple parameters. Results: In simulations, phantoms, and healthy subjects, the shortened optimized scan yielded similar myocardial T1 and T2 values as a previously described cardiac MRF technique. Impact: This study proposes a sequence optimization method for cardiac MRF with many potential applications, including designing shortened scans to reduce breathhold requirements and limit motion artifacts.