Automated Prediction of Postoperative Pancreatic Fistula Using Preoperative Computed Tomography

2026-05-29

AI Summary

Postoperative pancreatic fistula (POPF) is a serious complication following pancreatectomy, necessitating accurate preoperative risk assessment. This study proposes the first end-to-end deep learning framework that leverages preoperative CT imaging to automatically segment the pancreas and predict POPF risk, achieving full automation of the clinical workflow. The method integrates multiple 3D architectures—including a novel lightweight CNN3D, R(2+1)D ResNet-18, and ResNet-MC3-18—and demonstrates superior performance across multimodel evaluations. Beyond establishing a benchmark for pancreas-specific classification in CT imaging, this work provides an effective decision-support tool for preoperative risk stratification in clinical practice.

Abstract

Postoperative pancreatic fistula (POPF) is a serious complication after pancreatic resection, increasing morbidity, hospital stay, and healthcare costs. We present an automatic, end-to-end deep learning pipeline-from pancreatic segmentation to classification-for preoperative POPF risk estimation and stratification using preoperative CT scans. A data set with auto-segmented pancreas volumes and surgical outcomes was used to evaluate multiple architectures, including a custom lightweight 3D CNN baseline (CNN3D), R(2+1)D ResNet-18, and ResNet-MC3-18 models. Evaluation across multiple 3D architectures demonstrated promising predictive performance. This approach offers a clinically valuable tool and a methodological benchmark for pancreas-specific CT classification, supporting improved preoperative decision-making in pancreatic surgery.

Problem

• postoperative pancreatic fistula
• pancreatic resection
• preoperative CT
• risk prediction
• pancreatic surgery

Innovation

• postoperative pancreatic fistula
• deep learning
• 3D CT segmentation
• preoperative risk prediction
• end-to-end pipeline