Comparison between gradient and spin-echo (GRASE) and compressed sensing sequences for single breath-hold three-dimensional magnetic resonance cholangiopancreatography in patients with T1 hyperintense bile.

PURPOSE: To compare image quality and the detectability of gallstones in patients with T1 hyperintense bile between single breath-hold three-dimensional (3D) magnetic resonance cholangiopancreatography (MRCP) with gradient and spin-echo (GRASE) and with compressed sensing (CS).

METHODS

We retrospectively evaluated patients who underwent MRCP using GRASE and CS and had hyperintense bile on T1-weighted fat-suppressed images. The relative duct-to-periductal contrast ratios (RCs) of each bile duct segment were calculated. Pancreaticobiliary duct visibility, motion artifacts, background suppression, and overall image quality were scored on a 5-point scale. The Wilcoxon signed-rank test was used to analyze differences in qualitative and quantitative results. Diagnostic performance in detecting common bile duct (CBD) and gallbladder stones was assessed using receiver operating characteristic (ROC) curves.

RESULTS

In total, 96 patients were included in the study. RCs of all bile duct segments in GRASE MRCP were significantly lower than those in CS MRCP (p < 0.001). All biliary duct visibility and overall image quality had significantly higher scores in GRASE MRCP than in CS MRCP (p < 0.001-0.003). Area under ROC curves of GRASE MRCP and CS MRCP were 1.00 and 0.88 for CBD stone (p = 0.14) and 0.93 and 0.82 for gallbladder stone (p = 0.08), respectively.

CONCLUSIONS

GRASE MRCP provides better image quality than CS MRCP in patients with hyperintense bile on T1-weighted images. The detectability of biliary stones was also higher in GRASE MRCP than in CS MRCP, although not significantly.