A clinically validated 3D deep learning approach for quantifying vascular invasion in pancreatic cancer - Takeaways - MDSpire

A clinically validated 3D deep learning approach for quantifying vascular invasion in pancreatic cancer

  • By

  • Yajiao Zhang

  • Haoran Zhang

  • Yanzhao Yang

  • Chao Wu

  • Lei Zhang

  • Wei Xia

  • Xue Wang

  • Xiaohuan Zhang

  • Lixiu Cao

  • Manju Liu

  • Jing Zhang

  • Fuhua Yan

  • Baiyong Shen

  • Ning Wen

  • December 31, 2025

  • 0 min

Share

  • 1

    PAN-VIQ is an automated deep learning framework designed to assess vascular invasion in pancreatic ductal adenocarcinoma using contrast-enhanced CT scans.

  • 2

    The framework quantifies tumor-vessel interactions by segmenting pancreatic tumors and five critical vessels, enhancing preoperative evaluation accuracy.

  • 3

    PAN-VIQ was trained on 2130 cases and prospectively tested in 202 patients, achieving accuracies exceeding 90% in external validation.

  • 4

    The model outperformed junior radiologists and matched senior radiologists in accuracy and recall, indicating its potential to standardize assessments.

  • 5

    This deep learning approach addresses interobserver variability and provides a continuous, vessel-specific quantification of vascular invasion.

Original Source(s)

A clinically validated 3D deep learning approach for quantifying vascular invasion in pancreatic cancer

  • by Yajiao Zhang, Haoran Zhang, Yanzhao Yang, Chao Wu, Lei Zhang, Wei Xia, Xue Wang, Xiaohuan Zhang, Lixiu Cao, Manju Liu, Jing Zhang, Fuhua Yan, Baiyong Shen, Ning Wen

  • December 31, 2025

Related Content

Anatomically-guided Masked Autoencoder with Domain-Adaptive Prompting (AMAP) for multimodal cerebral aneurysm detection and segmentation

STD-Net: a spatio-temporal decoupling network for multiphasic liver lesion segmentation and characterization

Deep learning model for assessing survival benefits in hepatocellular carcinoma patients undergoing intra-arterial therapies based on proliferative subtype