Differentiation of mass-forming focal pancreatitis from pancreatic ductal adenocarcinoma: value of characterizing dynamic enhancement patterns on contrast-enhanced MR images by adding signal intensity color mapping

Eur Radiol. 2017 Apr;27(4):1722-1732. doi: 10.1007/s00330-016-4522-0. Epub 2016 Aug 10.

Abstract

Objectives: To evaluate the value of dynamic enhancement patterns on contrast-enhanced MR images by adding signal intensity colour mapping (SICM) to differentiate mass-forming focal pancreatitis (MFFP) from pancreatic ductal adenocarcinoma (PDAC).

Methods: Forty-one clinicopathologically proven MFFPs and 144 surgically confirmed PDACs were enrolled. Laboratory and MR imaging parameters were used to differentiate MFFP from PDAC. In particular, enhancement patterns on MR images adding SICM were evaluated. By using classification tree analysis (CTA), we determined the predictors for the differentiation of MFFP from PDAC.

Results: In the CTA, with all parameters except enhancement pattern on SICM images, ductal obstruction grade and T1 hypointensity grade of the pancreatic lesion were the first and second splitting predictor for differentiation of MFFP from PDAC, in order. By adding an enhancement pattern on the SICM images to CTA, the enhancement pattern was the only splitting predictor to differentiate MFFP from PDAC. The CTA model including enhancement pattern on SICM images has sensitivity of 78.0 %, specificity of 99.3 %, and accuracy of 94.6 % for differentiating MFFP from PDAC.

Conclusion: The characterization of enhancement pattern for pancreatic lesions on contrast-enhanced MR images adding SICM would be helpful to differentiate MFFP from PDAC.

Key points: • SICM was useful to characterize enhancement pattern. • Enhancement pattern on SICM was the only splitting predictor on CTA. • This model may be useful for differentiating MFFP from PDAC.

Keywords: Classification tree analysis; Magnetic resonance imaging; Mass-forming focal pancreatitis; Pancreatic ductal adenocarcinomas; Signal intensity colour mapping.

Publication types

  • Evaluation Study

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Carcinoma, Pancreatic Ductal / diagnostic imaging*
  • Color
  • Contrast Media
  • Diagnosis, Differential
  • Female
  • Humans
  • Image Enhancement
  • Image Interpretation, Computer-Assisted / methods
  • Magnetic Resonance Imaging / methods
  • Male
  • Middle Aged
  • Pancreatic Neoplasms / diagnostic imaging*
  • Pancreatitis / diagnostic imaging*
  • Retrospective Studies
  • Sensitivity and Specificity

Substances

  • Contrast Media