Numerical methods to achieve robust relaxometry mapping in multi-echo chemical shift-based MRI

We investigate on the feasibility of numerical methods to provide robust transverse relaxation R2∗ mapping in magnetic resonance imaging (MRI) applications by using latest theoretical models corrected for multiple confounding factors. Currently, a performance improvement in state-of-the-art MRI relaxometry algorithms is challenging because of a non-negligible bias and still unsolved numerical instabilities. Here, R2∗ mapping reconstructions, including complex-fitting with multi-spectral fat-correction using single-decay and double-decay formulation, are explored in order to identify optimal configuration parameters and performance limits. In addition results are evaluated by performing a comparison between single and multiple fat spectrum pre-calibration routines. Complex fitting and fat-correction with multi-exponential decay formulation outperforms the standard single-decay approximation in various diagnostic scenarios. In the study of neuromuscular disorders (NMD) our achievements aim also to highlight how the subdivision of image space into a number of partitioned areas and the adoption of multiple independent pre-calibration provides reduced fitting error if compared with single pre-calibration. The improvements are demonstrated in both simulations and in vivo applications. Together, such results suggest potential perspectives for the development of relaxometry as a reliable tool to improve tissue characterization and monitoring of NMD.