Accelerated Joint T1-T2 Fitting in Cardiac Magnetic Resonance Multitasking Using Extended Phase Graph Dictionaries

Motivation: Cardiac magnetic resonance 3D Multitasking can streamline clinical workflows through simultaneous T1-T2 fitting across the ventricle, but prolonged fitting times hinder widespread adoption of the technology. Goal(s): We aimed to determine if an extended phase graph (EPG) dictionary matching approach could reduce T1-T2 fitting times while maintaining measurement accuracy. Approach: We developed a numerical simulation to compare the original Bloch equation-based fitting model with the EPG dictionary matching approach against ground truth T1-T2 values. We subsequently compared the two fitting methods in-vivo. Results: The EPG dictionary matching significantly reduced T1-T2 fitting times without compromising the accuracy or precision of T1-T2 measurements. Impact: The significant reduction of T1-T2 fitting times through EPG dictionary matching improves the post-processing workflow of 3D Multitasking, potentially bringing the technology closer to clinical workflow integration.