[Comparison MR cholangiopancreatography with 3D-fast recovery fast spin echo in several different slice thicknesses].

PURPOSE To evaluate the technical quality and visibility of the biliary tree and pancreatic duct on magnetic resonance cholangiopancreatography (MRCP) images obtained with a single-breath-hold three-dimensional (3D) fast-recovery fast spin-echo (FRFSE) sequence in several different slice thicknesses. MATERIALS AND METHODS As a fundamental study, tubes of various inside diameters filled gadolinium solutions were acquired at 1.5 T in 3D-FRFSE. We observed error rate changes of volume inside the tubes and the visibility of thinner tubes. MRCP was performed at 1.5 T in 8 consecutive patients (4 men and 4 women, aged 22-58 years). Seven radiologists graded images obtained with each slice thickness in a blind fashion. Furthermore, we compared 1.4 mm slice thickness images with 1.8 mm slice thickness images in a continuous rating scale for the same patient. We assessed differences in technical quality, overall visibility, and six individual ductal segments of the biliary tree and pancreatic duct. RESULTS If slice thickness were thinner relative to diameter, the error rate would be closer to zero. But, when slice thickness was 0.8 mm, the error rate became clearly higher because of low intensity. In the fundamental study, we thought that the appropriate slice thickness is between 1.0 mm and 2.4 mm. The visibility of images of thinner tubes could be improved by having a thinner slice thickness. In particular, MRCP overall images generated from a 1.4 mm slice thickness were found to be significantly superior to those generated from a 1.8 mm slice thickness (p<0.001); this was also true as regards the pancreatic duct and cystic duct (p<0.01, p<0.05). CONCLUSION We conclude that a 1.4 mm slice thickness is appropriate for MRCP.