Real-world Evidence of Diffusion-weighted Imaging Combined With Magnetic Resonance Cholangiopancreatography for Pancreatic Tumor Screening: A Cross-Sectional Diagnostic Accuracy Study.

Abstract

The aim was to clarify the sensitivity and specificity of diffusion-weighted imaging, as well as of that in combination with magnetic resonance cholangiopancreatography for pancreatic tumor diagnosis in real-world clinical setting. Subjects were 217 consecutive patients who underwent both magnetic resonance imaging and contrast-enhanced ultrasound sonography. Cases positive for a pancreatic tumor were confirmed based on pathological diagnosis, whereas negative cases were defined when no solid pancreatic tumor was detected by contrast-enhanced ultrasound sonography or a solid mass was detected but the diagnosis was ultimately denied based on pathological results. Diffusion-weighted imaging-positive was defined as a case with high signals and magnetic resonance cholangiopancreatography-positive when localized main pancreatic duct stenosis with caudal dilation was detected.We calculated sensitivity and specificity of each modality and those in combination based on sequential use for pancreatic tumor diagnosis. Diffusion-weighted imaging showed a sensitivity of 94.4% and specificity of 94.5%, whereas those values for magnetic resonance cholangiopancreatography alone were 83.3% and 99.0%, respectively, and for the modalities in combination were 100% and 94.5%, respectively. Diffusion-weighted imaging was more sensitive than magnetic resonance cholangiopancreatography, whereas those used in combination resulted in increased sensitivity.