Multishot diffusion‐prepared magnitude‐stabilized balanced steady‐state free precession sequence for distortion‐free diffusion imaging

Abstract

To develop and evaluate a multishot diffusion-prepared (DP) magnitude-stabilized balanced steady-state free precession (bSSFP) diffusion imaging sequence with improved geometric fidelity. A signal spoiler (magnitude stabilizer; MS) was implemented in a DP-bSSFP diffusion sequence. Effects of magnitude stabilizers with respect to phase errors were simulated using Bloch simulation. Phantom study was conducted to compare the apparent diffusion coefficient (ADC) accuracy and geometric reliability, quantified using target registration error (TRE), with diffusion-weighted single-shot echo-planar imaging (DW-ssEPI). Six volunteers were recruited. DW-ssEPI, DP-bSSFP with and without ECG triggering, and DP-MS-bSSFP with and without ECG triggering were acquired 10 times with b = 500 s/mm2 in a single-shot manner to evaluate magnitude variability. Diffusion trace images and diffusion tensor images were acquired using a 4-shot DP-MS-bSSFP. Simulation showed that the DP-MS-bSSFP approach is insensitive to phase errors. The DP-MS-bSSFP approach had satisfactory ADC accuracy on the phantom with <5% difference with DW-ssEPI. The mean/max TRE for DW-ssEPI was 2.31/4.29 mm and was 0.51/1.20 mm for DP-MS-bSSFP. In the repeated single-shot study, DP-bSSFP without ECG triggering had severe signal void artifacts and exhibited a nonrepeatable pattern, which can be partially mitigated by ECG triggering. Adding the MS provided stable signal magnitude across all repetitions. High-quality ADC maps and color-coded fractional anisotropy maps were generated using the 4-shot DP-MS-bSSFP. The mean/max TRE was 2.89/10.80 mm for DW-ssEPI and 0.59/1.69 mm for DP-MS-bSSFP. Good agreements of white matter ADC, cerebrospinal fluid ADC, and white matter fractional anisotropy value were observed between DP-MS-bSSFP and DW-ssEPI. The proposed DP-MS-bSSFP approach provided high-quality diffusion-weighted and diffusion-tensor images with minimal geometric distortion.