In vitro and clinical studies of image acquisition in breath-hold MR cholangiopancreatography: Single-shot projection technique versus multislice technique

Abstract

OBJECTIVE. This study was undertaken to compare the in vitro and clinical value of two-dimensional multislice breath-hold MR cholangiopancreatography (MRCP) with a single-shot projection technique using a half-Fourier acquisition single-shot turbo spin-echo sequence. SUBJECTS AND METHODS. We examined 108 patients with pancreaticobiliary diseases, using breath-hold MRCP and a half-Fourier acquisition single-shot turbo spin-echo sequence on a 1.5-T MR unit with a body phased-array coil. Two data acquisition techniques were employed: multislice acquisition postprocessed by maximum intensity projection (MP) (multislice technique) and single-shot projection with a thick slice (projection technique). In the multislice technique, nine contiguous slices were obtained with a thickness of 5 mm (acquisition time, 18 sec). In the projection technique, a single slice was obtained with a thickness of 30, 50, or 70 mm (acquisition time, 2 sec). Contrast-to-noise ratio (CNR) between the common bile duct and the liver as well as detectability of normal structures and diseases were compared for these two acquisition techniques. In the multislice technique, source images were also evaluated. ERCP or percutaneous transhepatic cholangiography images were used as the gold standard. RESULTS. Most of the pancreatic duct and common bile duct was revealed on 54% and 100% of the projection images, respectively, and on 35% and 98% of the MIP images respectively. CNR was significantly higher with the multislice technique than with the projection technique (p < .01). With the projection technique, CNR decreased as slice thickness increased. Dilatation and occlusion of the pancreaticobiliary tree were equally well revealed by the two imaging techniques. However, abnormalities in the priampullary region and anomalies in the panceraticobiliary tree were more clearly seen on projection images than on MIP images (p < .05). Stones in the common bile duct, gallbladder, or intrahepatic bile duct were best seen on source images acquired by the multislice technique (83% sensitivity). Conclusion. Because of the absence of misregistration and the speed of image acquisition, breath-hold single-shot MRCP using the projection technique with a slice thickness of 30 or 50 mm consistently revealed the pancreaticobiliary tree and periampullary region with an acceptable CNR. Stones in the bile duct were best seen one the source images acquired by the MIP technique.