Enhancing physician support in pancreatic cancer diagnosis - New M-F-RCNN artificial intelligence model using endoscopic ultrasound

Sep 21, 2026

Uploaded by Skyler

Endoscopic ultrasound (EUS) plays a key role in early pancreatic cancer diagnosis, yet diagnostic errors can arise from limited image quality and operator experience. This study develops and validates a Modified Faster R-CNN (M-F-RCNN) deep-learning model that uses EUS images to locate cancerous regions. Training used EUS data from 155 patients across three centers (July 2022–July 2023), with external validation on 1,000 images and a reader study involving five EUS doctors.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages8

Reading Time20 min

AccessFree to download

Abstract

Background and study aims

Methods

Results

Conclusions

Introduction

Enhancing physician support in pancreatic cancer diagnosis: New M-F-RCNN artificial intelligence model using endoscopic ultrasound Authors Shan-shan Hu1, Bowen Duan2, Li Xu1, Danping Huang3, Xiaogang Liu1, Shihao Gou3, Xiaochen Zhao4, Jie Hou5, Shirong Tan5, lan ying He6, Ying Ye7, Xiaoli Xie8, Hong Shen1, Wei-hui Liu1 Institutions 1 Department of Gastroenterology and Hepatology, Sichuan Provincial People ʼ s Hospital, School of Medicine, University of Electronic Science and Technology of China, Chengdu, Sichuan, China 2 Endoscopy, Sichuan Cancer Hospital and Institute, Chengdu, China 3 Engineering and Science, Sichuan University of Science and Engineering Artificial Intelligence Key Laboratory of Sichuan Province, Yibin, China 4 Hepatobiliary Pancreatic Surgery, Sichuan Academy of Medical Sciences and Sichuan People's Hospital, Chengdu, China 5 Digestive Endoscopy Center of the Dongyuan, Sichuan Academy of Medical Sciences and Sichuan People's Hospital, Chengdu, China 6 Gastroenterology, Chongqing University Cancer Hospital, Chongqing, China 7 Department of Gastroenterology, Chengdu University of Traditional Chinese Medicine Affiliated Fifth People's hospital, Chengdu, China 8 Gastroenterology, The First People's Hospital of Longquanyi District Chengdu, Chengdu, China Key words Endoscopic ultrasonography, Pancreas, Endoscopy Upper GI Tract, Precancerous conditions & cancerous lesions (displasia and cancer) stomach, Diagnosis and imaging (inc chromoendoscopy, NBI, iSCAN, FICE, CLE) received 27.6.2024 accepted after revision 26.9.2024 accepted manuscript online 30.09.2024 Bibliography Endosc Int Open 2024; 12: E1277–E1284 DOI 10.1055/a-2422-9214 ISSN 2364-3722 © 2024. The Author(s) . This is an open access article published by Thieme under the terms of the Creative Commons Attribution-NonDerivative-NonCommercial License, permitting copying and reproduction so long as the original work is given appropriate credit. Contents may not be used for commercial purposes, or adapted, remixed, transformed or built upon. ([https://creativecommons.org/licenses/by-nc-nd/4.0/](https://creativecommons.org/licenses/by-nc-nd/4.0/)) Georg Thieme Verlag KG, Rüdigerstraße 14, 70469 Stuttgart, Germany Corresponding author Wei-hui Liu, Department of Gastroenterology and Hepatology, Sichuan Provincial People ʼs Hospital, School of Medicine, University of Electronic Science and Technology of China, 32 Western First Ring Road, 610072 Chengdu, Sichuan, China [audiliu12@163.com](audiliu12@163.com) ABSTRACT Background and study aims Endoscopic ultrasound (EUS) is vital for early pancreatic cancer diagnosis. Advances in artificial intelligence (AI), especially deep learning, have improved medical image analysis. We developed and validated the Modified Faster R-CNN (M-F-RCNN), an AI algorithm using EUS images to assist in diagnosing pancreatic cancer. Methods We collected EUS images from 155 patients across three endoscopy centers from July 2022 to July 2023. M-F-RCNN development involved enhancing feature information through data preprocessing and utilizing an improved Faster R-CNN model to identify cancerous regions. Its diagnostic capabilities were validated against an external set of 1,000 EUS images. In addition, five EUS doctors participated in a study comparing the M-F-RCNN model's performance with that of human experts, assessing diagnostic skill improvements with AI assistance. Results Internally, the M-F-RCNN model surpassed traditional algorithms with an average precision of 97.35%, accuracy of 96.49%, and recall rate of 5.44% . In external validation, its sensitivity, specificity, and accuracy were 91.7%, 91.5%, and 91 .6%, respectively, outperforming non-expert physicians. The model also significantly enhanced the diagnostic skills of doctors. Conclusions The M-F-RCNN model shows exceptional performance in diagnosing pancreatic cancer via EUS images, greatly improving diagnostic accurac

FAQs

What problem does the M-F-RCNN model target in pancreatic cancer diagnosis?

It targets diagnostic errors in pancreatic cancer detection using EUS images that can be influenced by image quality and physicians’ experience. The model assists by identifying cancerous regions from EUS imaging.

How was the M-F-RCNN model developed and validated?

The study collected EUS images from 155 patients across three endoscopy centers (July 2022–July 2023) and developed the model using preprocessing and an improved Faster R-CNN framework. Diagnostic performance was validated against an external set of 1,000 EUS images.

What were the main diagnostic performance results?

Internally, the model achieved average precision of 97.35%, accuracy of 96.49%, and a recall rate of 5.44%. In external validation, sensitivity, specificity, and accuracy were 91.7%, 91.5%, and 91.6%, respectively.

Sep 21, 2026

Uploaded by Skyler