Toward Brain Biosignature at 7 T-Repeatability of a Multiparametric CEST, QSM, QTI, and Sodium Imaging Protocol

To evaluate the repeatability of an extended multiparametric 7 T MRI protocol for voxel-wise tissue characterization combining morphological, CEST, QSM, QTI, and Sodium contrasts.

A repeatability study was performed in 12 volunteers without diagnosed neurological disease on a 7 T MRI system, using a comprehensive acquisition protocol with a combined acquisition time of 45:24 min, from which 45 contrast maps were derived. Quantitative repeatability was assessed using region-based and voxel-wise analyses. Precision and reliability were evaluated using the coefficient of variation (CV), repeatability coefficient (%RC), and intraclass correlation coefficient (ICC). Voxel-wise agreement between scan-rescan images was assessed using Pearson correlation coefficients.

Across contrasts spanning morphological, CEST, QSM, QTI, and ²³Na imaging, %RC was below 20% for 20 contrasts, with mean CV below 5% for most. Repeatability was highest for MP2RAGE T₁ maps and QTI MD (CV ≈ 1%-2%; ICC up to 0.86) and lowest for ²³Na imaging (ICC < 0.6). CEST contrasts were highly repeatable at low saturation power but more variable at higher power (CV up to ≈8%). QSM contrasts showed moderate ICC (0.46-0.82). Voxel-wise correlations exceeded r = 0.8 for most contrasts, confirming preserved spatial patterns across repeated scans.

Multiparametric 7 T MRI enables reproducible voxel-wise characterization of brain tissue across a wide range of contrasts. Despite variability in specific contrasts, the overall high repeatability and strong voxel-wise agreement support the feasibility of voxel-wise imaging approaches for quantitative brain imaging and provide a foundation for future applications in clinical research and neuroscience.