JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Differentiating small (< 2 cm) pancreatic ductal adenocarcinoma from neuroendocrine tumors with multiparametric MRI-based radiomic features

  • 0Department of Radiology, the Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, 310009, China.
European Radiology +

|

June 13, 2024

|

June 13, 2024

View abstract on PubMed

Summary

This summary is machine-generated.

Magnetic resonance imaging (MRI)-based radiomic analysis can differentiate small pancreatic ductal adenocarcinomas (PDACs) from pancreatic neuroendocrine tumors (PNETs). A fusion model combining radiomic, radiological, and clinical data demonstrated high accuracy in preoperative diagnosis.

Area Of Science

  • Radiology
  • Oncology
  • Medical Imaging

Background

  • Preoperative differentiation between small pancreatic ductal adenocarcinomas (PDACs) and pancreatic neuroendocrine tumors (PNETs) is clinically challenging.
  • Accurate diagnosis is crucial for appropriate treatment planning and patient management.

Purpose Of The Study

  • To evaluate the efficacy of multiparametric MRI-based radiomic analysis in distinguishing small PDACs from PNETs.
  • To develop and validate models for improved preoperative diagnosis.

Main Methods

  • Retrospective analysis of 197 patients (146 training, 51 validation) from two centers.
  • Extraction and selection of 7338 radiomic features from various MRI sequences.
  • Construction of radiomic score (Rad-score) and multivariable logistic regression models (radiological, radiomic, and fusion models).

Main Results

  • The radiomic model achieved high performance (AUCs 0.905-0.930).
  • The fusion model, integrating Rad-score with clinical and radiological features (CA19-9, tumor margins, pancreatic duct dilatation), showed superior performance (AUCs 0.977-0.941).
  • The fusion model demonstrated high sensitivity and specificity in both training and validation cohorts.

Conclusions

  • The MR-based Rad-score serves as a novel imaging biomarker for discriminating small PDACs from PNETs.
  • A fusion model incorporating radiomic, radiological, and clinical features offers a robust approach for the differential diagnosis of these small pancreatic tumors.
  • This approach can aid in preoperative differentiation, guiding clinical decision-making.

Keywords: