Kwangbeom Park, Jisup Lim, Seung Hwan Shin, Minkyeong Ryu, Hyungeun Shin, Minyoung Lee, Seung Wook Hong, Sung Wook Hwang, Sang Hyoung Park, Dong-Hoon Yang, Byong Duk Ye, Seung-Jae Myung, Suk-Kyun Yang, Namkug Kim, Jeong-Sik Byeon
{"title":"Artificial intelligence-aided colonoscopic differential diagnosis between Crohn's disease and gastrointestinal tuberculosis.","authors":"Kwangbeom Park, Jisup Lim, Seung Hwan Shin, Minkyeong Ryu, Hyungeun Shin, Minyoung Lee, Seung Wook Hong, Sung Wook Hwang, Sang Hyoung Park, Dong-Hoon Yang, Byong Duk Ye, Seung-Jae Myung, Suk-Kyun Yang, Namkug Kim, Jeong-Sik Byeon","doi":"10.1111/jgh.16788","DOIUrl":null,"url":null,"abstract":"<p><strong>Background and aim: </strong>Differentiating between Crohn's disease (CD) and gastrointestinal tuberculosis (GITB) is challenging. We aimed to evaluate the clinical applicability of an artificial intelligence (AI) model for this purpose.</p><p><strong>Methods: </strong>The AI model was developed and assessed using an internal dataset comprising 1,132 colonoscopy images of CD and 1,045 colonoscopy images of GITB at a tertiary referral center. Its stand-alone performance was further evaluated in an external dataset comprising 67 colonoscopy images of 17 CD patients and 63 colonoscopy images of 14 GITB patients from other institutions. Additionally, a crossover trial involving three expert endoscopists and three trainee endoscopists compared AI-assisted and unassisted human interpretations.</p><p><strong>Results: </strong>In the internal dataset, the sensitivity, specificity, and accuracy of the AI model in distinguishing between CD and GITB were 95.3%, 100.0%, and 97.7%, respectively, with an area under the ROC curve of 0.997. In the external dataset, the AI model exhibited a sensitivity, specificity, and accuracy of 77.8%, 85.1%, and 81.5%, respectively, with an area under the ROC curve of 0.877. In the human endoscopist trial, AI assistance increased the pooled accuracy of the six endoscopists from 86.2% to 88.8% (P = 0.010). While AI did not significantly enhance diagnostic accuracy for the experts (96.7% with AI vs 95.6% without, P = 0.360), it significantly improved accuracy for the trainees (81.0% vs 76.7%, P = 0.002).</p><p><strong>Conclusions: </strong>This AI model shows potential in aiding the accurate differential diagnosis between CD and GITB, particularly benefiting less experienced endoscopists.</p>","PeriodicalId":15877,"journal":{"name":"Journal of Gastroenterology and Hepatology","volume":" ","pages":""},"PeriodicalIF":3.7000,"publicationDate":"2024-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Gastroenterology and Hepatology","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1111/jgh.16788","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"GASTROENTEROLOGY & HEPATOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Background and aim: Differentiating between Crohn's disease (CD) and gastrointestinal tuberculosis (GITB) is challenging. We aimed to evaluate the clinical applicability of an artificial intelligence (AI) model for this purpose.
Methods: The AI model was developed and assessed using an internal dataset comprising 1,132 colonoscopy images of CD and 1,045 colonoscopy images of GITB at a tertiary referral center. Its stand-alone performance was further evaluated in an external dataset comprising 67 colonoscopy images of 17 CD patients and 63 colonoscopy images of 14 GITB patients from other institutions. Additionally, a crossover trial involving three expert endoscopists and three trainee endoscopists compared AI-assisted and unassisted human interpretations.
Results: In the internal dataset, the sensitivity, specificity, and accuracy of the AI model in distinguishing between CD and GITB were 95.3%, 100.0%, and 97.7%, respectively, with an area under the ROC curve of 0.997. In the external dataset, the AI model exhibited a sensitivity, specificity, and accuracy of 77.8%, 85.1%, and 81.5%, respectively, with an area under the ROC curve of 0.877. In the human endoscopist trial, AI assistance increased the pooled accuracy of the six endoscopists from 86.2% to 88.8% (P = 0.010). While AI did not significantly enhance diagnostic accuracy for the experts (96.7% with AI vs 95.6% without, P = 0.360), it significantly improved accuracy for the trainees (81.0% vs 76.7%, P = 0.002).
Conclusions: This AI model shows potential in aiding the accurate differential diagnosis between CD and GITB, particularly benefiting less experienced endoscopists.
期刊介绍:
Journal of Gastroenterology and Hepatology is produced 12 times per year and publishes peer-reviewed original papers, reviews and editorials concerned with clinical practice and research in the fields of hepatology, gastroenterology and endoscopy. Papers cover the medical, radiological, pathological, biochemical, physiological and historical aspects of the subject areas. All submitted papers are reviewed by at least two referees expert in the field of the submitted paper.