Rapid and simplified post-processing for simultaneous B0 and B1 mapping in the application of CEST

PURPOSE: B 0 and B 1 inhomogeneity corrections are crucial for accurate CEST imaging, particularly at ultra-high-field MRI. WASABI provides high fidelity δ ω  delta omega and B 1 {mathrm{B}}_1 maps but suffers from prolonged post-processing and sensitivity to local minima. Our objective was to design an alternative to WASABI’s Levenberg-Marquardt-based optimization approach to improve both post-processing speed and accuracy.

METHODS: A direct relationship was derived between δ ω  delta omega and B 1  {mathrm{B}}_1 $$ values and information contained in WASABI Z-spectra. Seven in vivo brain datasets were acquired at 7T.

RESULTS: The proposed approach, called RAbi DIstance SearcH (RADISH), accelerated post-processing by two orders of magnitude, with improved estimation across all brain slices. Maps produced with RADISH were consistent with those produced by unartifacted areas in the original approach, with agreements within 1 Hz and 0.5% for δ ω  delta omega  and r B 1  {mathrm{B}}_1 maps, respectively. The percentage of whole-head artifacts was reduced from 3.90% to 1.05%.

CONCLUSIONS: Improvement in speed and robustness provided by RADISH allows for reliable generation of δ ω  delta omega $$ and B 1  {B}_1 maps, contributing to making quantitative CEST imaging at ultra-high-field more reliable and advancing its clinical feasibility.