A logistic regression model developed from various risk factors effectively predicts early esophageal cancer with submucosal infiltration beyond 200 micrometers. Key predictive factors include endoscopic findings, esophageal wall thickening, pickled food intake, platelet-lymphocyte ratio, tumor size, circumferential mucosal defect percentage, and preoperative pathological type. The model demonstrated high efficacy, achieving an area under the curve of 0.922 in training and 0.899 in validation, underscoring its potential for clinical application.
Journal Article by Cui Y, Luo Z (…) Wang X et 7 al. in BMC Gastroenterol
© 2024. The Author(s).