Rapid and quantitative CEST-MRI sequence using water presaturation

Despite the significant potential for in vivo metabolic imaging in preclinical and clinical applications, CEST MRI suffers from long scan time and inaccurate quantification. This study aims to suppress the contaminations among signals under different frequencies, which could shorten the TR and thereby facilitate CEST imaging acceleration and quantification.

A novel sequence is proposed by applying a water-presaturation (WPS) module at the beginning of each TR. WPS CEST quickly knocks down the residual signal from previous TRs so that the magnetization of all TRs recovers from zero, which aligns well with the formula of quasi-steady-state theorem and enables accurate quantification within shorter TR. WPS CEST was assessed by simulations, creatine phantom, and healthy human brain scans at 3 T.

In simulation and phantom experiment, WPS CEST allows accurate estimation of exchange rate (k_sw) using omega plot and using shorter delay time (T_d) and saturation time (T_s) (e.g., 1 s/1 s) compared with the conventional CEST. Simulations further showed that WPS CEST could obtain consistent spin-lock relaxation (R_1ρ) values over varied T_ds and T_ss. Six human scans indicated that R_1ρ collected from conventional sequences showed significant differences between two groups with T_d and T_s of (1 s/1 s) and (2 s/2 s) (amide: 1.721 ± 0.051 s^-1 vs. 1.622 ± 0.050 s^-1, p = 0.001; nuclear Overhauser enhancement: 1.792 ± 0.046 s^-1 vs. 1.687 ± 0.053 s^-1, p = 0.004), whereas WPS CEST scans using these 2 T_d/T_s values obtained the same mean R_1ρ (amide: 1.616 ± 0.053 s^-1 vs. 1.616 ± 0.048 s^-1, p = 0.862; nuclear Overhauser enhancement: 1.688 ± 0.064 s^-1 vs. 1.684 ± 0.054 s^-1, p = 0.544).

WPS CEST demonstrated accurate quantitation within shorter TR compared with conventional sequences, and thereby may allow rapid quantitative CEST scans in various situations.