Evaluating the effectiveness of distortion self-correction for CEST-EPI.

Purpose: EPI is a fast acquisition sequence, but suffers from geometric distortion because of B field inhomogeneity. This study aims to evaluate the effectiveness of using ΔB map generated from single-shot CEST-EPI to achieve distortion self-correction (DISC).

Methods: CEST MRI usually requires B correction during postprocessing, and the ΔB map can be calculated directly from Z-spectra without extra scan. We propose to use the ΔB map to correct the geometry distortion induced by B field inhomogeneity in CEST-EPI. The effectiveness of DISC strategy for CEST-EPI was evaluated on a creatine phantom, healthy and tumor mice at 3 T, and AQP4 heterozygotes mice at 11.7 T. For both the original CEST images and the generated CEST contrast maps, the spatial improvement was first confirmed by the visual comparison and then quantitatively assessed by the structural similarity index measure (SSIM) comparison together with correlation analysis.

Results: DISC-CEST-EPI showed higher SSIM and spatial consistency compared to CEST-EPI when using CEST-rapid acquisition with relaxation enhancement as a reference in vitro and in vivo at 3 T. For in vivo experiments at 11.7 T, SSIM values of amide proton transfer and relayed nuclear Overhauser effects maps of DISC-CEST-EPI were slightly higher than CEST-EPI and comparable with field-mapping and top-up.

Conclusion: DISC-CEST-EPI can correct the distortion in CEST-EPI at 3 T, leading to improved SSIM and CEST quantification. This approach has the potential to enhance image quality and improve diagnostic accuracy for single-shot CEST-EPI at especially low-field MRI. However, the performance of DISC-CEST-EPI is limited at 11.7 T, and further advancements are necessary to enhance its functionality at ultra-high fields.