Development of intravoxel inhomogeneity correction forchemical exchange saturation transfer spectral imaging: a high-resolution field map-based deconvolution algorithm for magnetic field inhomogeneity correction.

Abstract

CEST MRI is sensitive to dilute proteins/peptides and microenvironmental properties yet susceptible to magnetic field inhomogeneity. We aimed to develop a high-resolution field map-basedCEST intravoxel inhomogeneity correction (CIVIC) algorithm for CEST Z-spectral imaging. The proposed CIVIC approach treats the intravoxel inhomogeneity as a point spread function and applies the deconvolution algorithm to reconstruct the original Z-spectrum. We simulated the effect of B0 field inhomogeneity on CEST measurement and tested the efficacy of the proposed CIVIC algorithm. We also performed CEST MRI on a dual-pH Creatine-gel phantom under varied field homogeneity conditions and compared the CEST MRI contrast-to-noise ratio from the raw Z-spectrum, water saturation shift referencing, and the proposed CIVIC methods. The numerical simulation showed that the CIVIC algorithm remains effective even in the case of symmetric field dispersion with a 0 mean shift. The experimental results confirmed that the proposed CIVIC method substantially improves the CEST MRI contrast-to-noise ratio under different field homogeneity conditions. Our study established a new intravoxel B0 inhomogeneity correction algorithm, promising to facilitate CEST spectral imaging in challenging experimental conditions.