Reproducibility of rapid multi-parameter mapping at 3T and 7T with highly segmented and accelerated 3D-EPI

Quantitative multi-parameter mapping (MPM) has been shown to provide good longitudinal and cross-sectional reproducibility for clinical research. Unfortunately, acquisition times (TAs) are typically infeasible for routine scanning at high resolutions.

A fast whole-brain MPM protocol based on interleaved multi-shot 3D-EPI with controlled aliasing (SC-EPI) at 3T and 7T is proposed and compared with MPM using a standard spoiled gradient echo (FLASH) sequence. Four parameters (R₁ , PD, $$ , and MTsat) were measured in less than 3 min at 1 mm isotropic resolution. Five subjects went through the same scanning sessions twice at each scanner. The intra-subject coefficient of variation (scan-rescan) (CoV) was estimated for each protocol and scanner to assess the longitudinal reproducibility.

At 3T, the CoV of SC-EPI ranged between 1.2%-4.8% for PD and R₁ , 2.8%-10.6% for $$ and MTsat, which was comparable with FLASH (0.6%-4.9% for PD and R₁ , 2.6%-11.3% for $$ and MTsat). At 7T, where the SC-EPI TA was reduced to ∼2 min, the CoV of SC-EPI (1.4%-10.6% for PD, R₁ , and $$ ) was 1.2-2.4 times larger than the CoV of FLASH (1.0%-15%) and MTsat showed much higher variability across subjects. The SC-EPI-MPM protocol at 3T showed high reproducibility and yielded stable quantitative maps at a clinically feasible resolution and scan time, whereas at 7T, MT saturation homogeneity needs to be improved.

SC-EPI-based MPM is feasible as an additional MRI modality in clinical or population studies where the parameters offer great potential as biomarkers.