A convolutional neural network-based system for identifying neuroendocrine neoplasms and multiple types of lesions in the pancreas using EUS (with videos).
Jie-Kun Ni, Ze-Le Ling, Xiao Liang, Yi-Hao Song, Guo-Ming Zhang, Chang-Xu Chen, Li-Mei Wang, Peng Wang, Guang-Chao Li, Shi-Yang Ma, Jun Gao, Le Chang, Xin-Xin Zhang, Ning Zhong, Zhen Li
{"title":"A convolutional neural network-based system for identifying neuroendocrine neoplasms and multiple types of lesions in the pancreas using EUS (with videos).","authors":"Jie-Kun Ni, Ze-Le Ling, Xiao Liang, Yi-Hao Song, Guo-Ming Zhang, Chang-Xu Chen, Li-Mei Wang, Peng Wang, Guang-Chao Li, Shi-Yang Ma, Jun Gao, Le Chang, Xin-Xin Zhang, Ning Zhong, Zhen Li","doi":"10.1016/j.gie.2024.10.013","DOIUrl":null,"url":null,"abstract":"<p><strong>Background and aims: </strong>EUS is sensitive in detecting pancreatic neuroendocrine neoplasm (pNEN). However, the endoscopic diagnosis of pNEN is operator-dependent and time-consuming because pNEN mimics normal pancreas and other pancreatic lesions. We intended to develop a convolutional neural network (CNN)-based system, named iEUS, for identifying pNEN and multiple types of pancreatic lesions using EUS.</p><p><strong>Methods: </strong>Retrospective data of 12,200 EUS images obtained from pNEN and non-pNEN pancreatic lesions, including pancreatic ductal adenocarcinoma (PDAC), autoimmune pancreatitis (AIP), and pancreatic cystic neoplasm (PCN), were used to develop iEUS, which was composed of a 2-category (pNEN or non-pNEN pancreatic lesions) classification model (CNN1) and a 4-category (pNEN, PDAC, AIP, or PCN) classification model (CNN2). Videos from consecutive patients were prospectively collected for a human-iEUS contest to evaluate the performance of iEUS.</p><p><strong>Results: </strong>Five hundred seventy-three patients were enrolled in this study. In the human-iEUS contest containing 203 videos, CNN1 and CNN2 showed an accuracy of 84.2% and 88.2% for diagnosing pNEN, respectively, which were significantly higher than that of novices (75.4%) and comparable with intermediate endosonographers (85.5%) and experts (85.5%). In addition, CNN2 showed an accuracy of 86.2%, 97.0%, and 97.0% for diagnosing PDAC, AIP, and PCN, respectively. With the assistance of iEUS, the sensitivity of endosonographers at all 3 levels in diagnosing pNEN has significantly improved (64.6% vs 44.8%, 87.5% vs 71.9%, and 74.0% vs 57.6%, respectively).</p><p><strong>Conclusions: </strong>The iEUS precisely diagnosed pNEN and other confusing pancreatic lesions and thus can assist endosonographers in achieving more accessible and accurate endoscopic diagnoses with EUS. (Clinical trial registration number: ChiCTR2100049697.).</p>","PeriodicalId":12542,"journal":{"name":"Gastrointestinal endoscopy","volume":" ","pages":""},"PeriodicalIF":6.7000,"publicationDate":"2024-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Gastrointestinal endoscopy","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1016/j.gie.2024.10.013","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"GASTROENTEROLOGY & HEPATOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Background and aims: EUS is sensitive in detecting pancreatic neuroendocrine neoplasm (pNEN). However, the endoscopic diagnosis of pNEN is operator-dependent and time-consuming because pNEN mimics normal pancreas and other pancreatic lesions. We intended to develop a convolutional neural network (CNN)-based system, named iEUS, for identifying pNEN and multiple types of pancreatic lesions using EUS.
Methods: Retrospective data of 12,200 EUS images obtained from pNEN and non-pNEN pancreatic lesions, including pancreatic ductal adenocarcinoma (PDAC), autoimmune pancreatitis (AIP), and pancreatic cystic neoplasm (PCN), were used to develop iEUS, which was composed of a 2-category (pNEN or non-pNEN pancreatic lesions) classification model (CNN1) and a 4-category (pNEN, PDAC, AIP, or PCN) classification model (CNN2). Videos from consecutive patients were prospectively collected for a human-iEUS contest to evaluate the performance of iEUS.
Results: Five hundred seventy-three patients were enrolled in this study. In the human-iEUS contest containing 203 videos, CNN1 and CNN2 showed an accuracy of 84.2% and 88.2% for diagnosing pNEN, respectively, which were significantly higher than that of novices (75.4%) and comparable with intermediate endosonographers (85.5%) and experts (85.5%). In addition, CNN2 showed an accuracy of 86.2%, 97.0%, and 97.0% for diagnosing PDAC, AIP, and PCN, respectively. With the assistance of iEUS, the sensitivity of endosonographers at all 3 levels in diagnosing pNEN has significantly improved (64.6% vs 44.8%, 87.5% vs 71.9%, and 74.0% vs 57.6%, respectively).
Conclusions: The iEUS precisely diagnosed pNEN and other confusing pancreatic lesions and thus can assist endosonographers in achieving more accessible and accurate endoscopic diagnoses with EUS. (Clinical trial registration number: ChiCTR2100049697.).
期刊介绍:
Gastrointestinal Endoscopy is a journal publishing original, peer-reviewed articles on endoscopic procedures for studying, diagnosing, and treating digestive diseases. It covers outcomes research, prospective studies, and controlled trials of new endoscopic instruments and treatment methods. The online features include full-text articles, video and audio clips, and MEDLINE links. The journal serves as an international forum for the latest developments in the specialty, offering challenging reports from authorities worldwide. It also publishes abstracts of significant articles from other clinical publications, accompanied by expert commentaries.