Compressed-Sensing Accelerated 3-Dimensional Magnetic Resonance Cholangiopancreatography Application in Suspected Pancreatic Diseases

Objectives: The aims of this study were to prospectively evaluate image quality, duct visibility, and diagnostic performance in duct-related pathologies of compressed-sensing (CS) accelerated 3-dimensional (3D) magnetic resonance cholangiopancreatography (MRCP) prototype protocols and compare these with those of conventional 3D MRCP protocol in patients with suspected pancreatic diseases. Material and Methods: The institutional review board approved this prospective study and all patients provided written informed consent. A total of 80 patients (47 men and 33 women; median age, 57 years; age range, 24-87 years) underwent 3D MRCP at 3.0 T. Three protocols were performed in each patient in random order: CS breath-hold (BH) protocol, CS navigator-triggered (NT) protocol, and conventional NT protocol. The acquisition time of each protocol was recorded. Image quality and duct visibility were independently rated in random order on a 5-point scale by 2 radiologists, who were blinded to the protocols. Receiver operating characteristic curves were generated, and area under the curve (A(z) value) was used to compare the diagnostic performance of each protocol in duct-related pathologies. Results: Acquisition time was 17 seconds for the CS-BH and 134.1 +/- 33.5 seconds for the CS-NT protocol, both being significantly shorter than the conventional NT protocol (364.7 +/- 78.4 seconds; both P < 0.01). The CS-BH MRCP protocol showed significantly less artifacts compared with the CS-NT and conventional NT protocols (both P < 0.01). Visualization of bile ducts was comparable in all 3 protocols, whereas CS-NT and conventional NT MRCP depicted pancreatic duct better than CS-BH MRCP did (for proximal, middle, and distal segment; all P < 0.05). Compressed-sensing-NT MRCP had the highest diagnostic performance for detecting ductal anomalies, long-segment duct stenosis, abnormal branch ducts, and communication between cystic lesion and pancreatic duct (mean A(z) value, 0.943-0.983). Conclusions: Compressed-sensing MRCP is feasible in patients with suspected pancreatic diseases. Compressed-sensing-NT MRCP demonstrated superior diagnostic accuracy for duct-related pathologies.