Enhancing the Predictions of Cytomegalovirus Infection in Severe Ulcerative Colitis Using a Deep Learning Ensemble Model: Development and Validation Study

• Published in: JMIR medical informatics Vol. 13; p. e64987
• Main Authors: Kim, Jeong Heon, Choe, A Reum, Byeon, Ju Ran, Park, Yehyun, Song, Eun Mi, Kim, Seong-Eun, Jeong, Eui Sun, Lee, Rena, Kim, Jin Sung, Ahn, So Hyun, Jung, Sung Ae
• Format: Journal Article
• Language: English
• Published: Canada JMIR Publications 01.07.2025
• Subjects: Accuracy; Artificial intelligence; Automation; Biopsy; Blood tests; Clinical Information and Decision Making; Clinical outcomes; Colitis, Ulcerative - complications; Colitis, Ulcerative - virology; Colorectal surgery; Cytomegalovirus; Cytomegalovirus Infections - diagnosis; Datasets; Decision Support for Health Professionals; Deep Learning; Endoscopy; Female; Gastroenterology and Hepatology; Herpes viruses; Humans; Imaging Informatics; Infections; Infectious Diseases (non-STD/STI); Inflammatory bowel disease; Intervention; Machine Learning; Male; Medical imaging; Morphology; Ontologies, Classifications, and Coding; Original Paper; Patients; Validation studies; Viral infections; deep learning; endoscopy; ulcerative colitis; cytomegalovirus; classification

Cytomegalovirus (CMV) reactivation in patients with severe ulcerative colitis (UC) leads to worse outcomes; yet, early detection remains challenging due to the reliance on time-intensive biopsy procedures. This study explores the use of deep learning to differentiate CMV from severe UC through endos...