Related Experiment Video
Updated: Oct 17, 2026
Robotic Duodenum-preserving Total Pancreatic Head Resection for Intraductal Papillary Mucinous Neoplasms
Published on: April 17, 2026
Quantitative CT-Derived Imaging Features Improve Preoperative Risk Stratification of Intraductal Papillary Mucinous
Tae Seung Lee1, Hong Min Shin1, In-Ho Park1
1Department of Internal Medicine and Liver Research Institute, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Republic of Korea.
Background:
Accurate preoperative identification of high-grade dysplasia (HGD) or malignancy in intraductal papillary mucinous neoplasms (IPMNs) remains challenging using conventional guideline-based imaging criteria. Quantitative computed tomography (CT)-based features may detect subtle structural changes preceding overt radiologic findings. This study aimed to evaluate whether the pancreatic index (PI) and additional three-dimensional CT-derived quantitative features improve prediction of HGD or malignancy in IPMNs.
Methods:
This single-center retrospective study included patients who underwent surgical resection for pathologically confirmed IPMNs between January 2001 and August 2024 and had available preoperative pancreas-protocol CT. Three-dimensional quantitative parameters were analyzed.
Results:
Among 333 patients, 204 had low-grade dysplasia (LGD), 56 had HGD, and 73 had malignancy. PI decreased with increasing histologic grade (75.6 ± 10.2% in LGD, 74.5 ± 12.0% in HGD, 70.5 ± 12.6% in malignancy; p = 0.003). Pancreatic Hounsfield unit standard deviation (HU-SD) was significantly higher in HGD and malignancy than in LGD (25.6 ± 5.9, 29.0 ± 8.7, and 29.6 ± 6.2, respectively; p < 0.001). A combined model integrating guideline-based criteria with PI, CI, cystic HU-SD, pancreatic HU-SD, and pancreatic duct volume demonstrated better discrimination than the guideline-only model (area under the curve 0.800 vs. 0.760; p = 0.007).
Conclusions:
Quantitative CT-derived imaging features provide incremental value for predicting HGD and malignancy when integrated with conventional guideline-based criteria. These noninvasive imaging biomarkers may enhance risk stratification and provide additional information to support clinical decision-making when integrated with conventional criteria. Further external validation is required to establish their clinical utility in routine practice.