{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,18]],"date-time":"2026-10-18T16:41:17Z","timestamp":1792341677127,"version":"4.4.0"},"reference-count":10,"publisher":"Ovid Technologies (Wolters Kluwer Health)","content-domain":{"domain":["lww.com","ovid.com"],"crossmark-restriction":true},"short-container-title":[],"abstract":"Abstract<\/jats:title>\n \n Background:<\/jats:title>\n Gastritis diagnosis via endoscopy remains highly subjective, often leading to inter-observer variability and the misdiagnosis of early inflammatory changes as congestive portal gastropathy. While Computer-Aided Diagnosis (CAD) shows promise, the \u201cpaucity of labelled data\u201d remains a significant barrier to clinical implementation.<\/jats:p>\n <\/jats:sec>\n \n Aim and Objectives:<\/jats:title>\n This study aimed to develop and validate a scalable deep-learning solution using a hybrid CNN-Vision Transformer (ViT) architecture to assist endoscopists in the real-time severity stratification of gastritis (Normal, Mild, Moderate, and Severe).<\/jats:p>\n <\/jats:sec>\n \n Materials and Methods:<\/jats:title>\n We utilized 4,462 high-definition, anonymised endoscopic images annotated by expert gastroenterologists using the Kyoto Classification as a benchmark. A multi-branch parallel processing architecture was developed, combining a CNN-based branch for local feature extraction (texture and erythema) and a Vision Transformer (ViT) for global context learning. Data augmentation expanded the training set to over 68,000 images to prevent overfitting. Performance was evaluated using Accuracy, Precision, Recall, and the Matthews Correlation Coefficient (MCC).<\/jats:p>\n <\/jats:sec>\n \n Results:<\/jats:title>\n The hybrid model achieved an Overall Classification Accuracy of 91.4% and a Matthews Correlation Coefficient (MCC) of 0.90, indicating a near-perfect correlation with clinical ground truth. The system demonstrated exceptional performance in the \u201cNormal\u201d category with 99% specificity, providing a highly dependable negative result. Precision (0.8892) and Recall (0.8481) remained robust across all categories, significantly reducing the risk of over-diagnosing gastropathy as gastritis.<\/jats:p>\n <\/jats:sec>\n \n Conclusion:<\/jats:title>\n The proposed hybrid AI architecture provides a robust, real-time \u201csecond opinion\u201d tool for endoscopists. By integrating local and global mucosal patterns, the system achieves high-precision severity stratification, potentially reducing unnecessary biopsies and minimizing diagnostic errors caused by inter-observer variability in the endoscopy suite.<\/jats:p>\n <\/jats:sec>","DOI":"10.4103\/jmas.jmas_224_26","type":"journal-article","created":{"date-parts":[[2026,10,18]],"date-time":"2026-10-18T16:00:08Z","timestamp":1792339208000},"update-policy":"https:\/\/doi.org\/10.1097\/lww.0000000000001000","source":"Crossref","is-referenced-by-count":0,"title":["Evaluation of an artificial intelligence model in assessing the severity of gastritis during gastroscopy: A pilot study"],"prefix":"10.4103","author":[{"given":"Krishnamoorthy","family":"Rajaram","sequence":"first","affiliation":[{"name":"Department of Electronics and Communication Engineering, Velalar College of Engineering and Technology, Thindal, Erode, Tamil Nadu, India"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Palaniappan","family":"Sakthivel","sequence":"additional","affiliation":[{"name":"Department of Electrical and Electronics Engineering, Velalar College of Engineering and Technology, Thindal, Erode, Tamil Nadu, India"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sundaram","family":"Easwaramoorthy","sequence":"additional","affiliation":[{"name":"Department of Minimal Access Surgery, Lotus Hospital and Research Centre, Erode, Tamil Nadu, India"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chandrasekar","family":"Sakthivel","sequence":"additional","affiliation":[{"name":"Department of Surgery, Lotus Hospital, Erode, Tamil Nadu, India"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"276","published-online":{"date-parts":[[2026,8,26]]},"reference":[{"key":"R1-20261018","doi-asserted-by":"crossref","first-page":"1532640","DOI":"10.3389\/fmed.2025.1532640","article-title":"Artificial intelligence in endoscopy and colonoscopy: A comprehensive bibliometric analysis of global research trends","volume":"12","author":"Letafatkar","year":"2025","journal-title":"Front Med (Lausanne)"},{"key":"R2-20261018","first-page":"1","article-title":"Kyoto classification of gastritis, virtual chromoendoscopy and artificial intelligence","volume":"15","author":"Kamada","year":"2023","journal-title":"World J Gastrointest Endosc"},{"key":"R3-20261018","doi-asserted-by":"crossref","first-page":"e33942","DOI":"10.1097\/MD.0000000000033942","article-title":"The significance of endoscopic Kyoto classification of gastritis in the gastric cancer risk assessment: A systematic review and metaanalysis","volume":"102","author":"Zhang","year":"2023","journal-title":"Medicine (Baltimore)"},{"key":"R4-20261018","doi-asserted-by":"crossref","first-page":"579","DOI":"10.2169\/internalmedicine.56.7775","article-title":"Efficacy of the Kyoto classification of gastritis in identifying patients at high risk for gastric cancer","volume":"56","author":"Sugimoto","year":"2017","journal-title":"Intern Med"},{"key":"R5-20261018","doi-asserted-by":"crossref","first-page":"6491","DOI":"10.3748\/wjg.v28.i43.6078","article-title":"Kyoto classification of gastritis: Advances and future prospects","volume":"28","author":"Toyoshima","year":"2022","journal-title":"World J Gastroenterol"},{"key":"R6-20261018","first-page":"12","article-title":"The paucity of labelled data in medical AI: Strategies for clinical validation","volume":"7","author":"Dosanjh","year":"2024","journal-title":"J Med Artif Intell"},{"key":"R7-20261018","first-page":"1560","article-title":"Vision transformers versus convolutional neural networks in medical imaging: A comparative study of global and local feature extraction","volume":"42","author":"Lai","year":"2023","journal-title":"IEEE Trans Med Imaging"},{"key":"R8-20261018","doi-asserted-by":"crossref","first-page":"1122247","DOI":"10.3389\/fonc.2023.1122247","article-title":"Deep learning-assisted diagnosis of chronic atrophic gastritis in endoscopy","volume":"13","author":"Shi","year":"2023","journal-title":"Front Oncol"},{"key":"R9-20261018","first-page":"1082","article-title":"Application of artificial intelligence using a convolutional neural network for diagnosis of early gastric cancer based on magnifying endoscopy with narrow-band imaging","volume":"93","author":"Ueyama","year":"2021","journal-title":"Gastrointest Endosc"},{"key":"R10-20261018","first-page":"87","article-title":"An endoscopic recognition of the atrophic border and its significance in chronic gastritis","volume":"3","author":"Kimura","year":"1966","journal-title":"Endoscopy"}],"container-title":["Journal of Minimal Access Surgery"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.ovid.com\/10.4103\/jmas.jmas_224_26","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,18]],"date-time":"2026-10-18T16:00:10Z","timestamp":1792339210000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ovid.com\/10.4103\/jmas.jmas_224_26"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,26]]},"references-count":10,"URL":"https:\/\/doi.org\/10.4103\/jmas.jmas_224_26","relation":{},"ISSN":["0972-9941","1998-3921"],"issn-type":[{"value":"0972-9941","type":"print"},{"value":"1998-3921","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,26]]},"assertion":[{"value":"2026-04-15","name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-07-14","name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}}]}}