Improvement of water saturation shift referencing by sequence and analysis optimization to enhance chemical exchange saturation transfer imaging.

PURPOSE: To optimize B0-field inhomogeneity correction for chemical exchange saturation transfer (CEST) imaging by investigating different water saturation shift referencing (WASSR) Z-spectrum shapes and different frequency correction techniques. METHODS: WASSR Z-spectra were simulated for different B1-fields and pulse durations (PD). Two parameter settings were used for further simulations and experiments (WASSR1: B1=0.1 µT, PD=50ms; WASSR2: B1=0.3 µT, PD=40ms). Four frequency correction techniques were investigated: 1) MinW: Minimum of the spline-interpolated WASSR-spectrum; 2) MSCF: maximum symmetry center frequency algorithm; 3) PMSCF: further development of MSCF algorithm; 4) BFit: fit with Bloch equations. Performance of frequency correction was assessed with Monte-Carlo simulations and in-vivo MR examinations in the brain and intervertebral disks. RESULTS: Different shapes of WASSR-Z-spectra were obtained by changing B1 and PD including spectra with one (1-Peak) or two (2-Peak) minima. WASSR1 resulted in 1-Peak WASSR-spectrum, whereas WASSR2 resulted in 2-Peak WASSR-spectrum. Both Monte-Carlo simulations and in-vivo MR examinations revealed highest accuracy of field-inhomogeneity correction with WASSR1 combined with PMSCF or BFit. CONCLUSION: Using a WASSR sequence, which results in a Z-spectrum with a single absorption peak, in combination with advanced postprocessing algorithms enables improved B0-field inhomogeneity correction for CEST imaging.