Machine learning model predicts malignant potential of gastric tumors

A novel machine-learning model effectively predicts the malignant potential of gastric gastrointestinal stromal tumors (GISTs), identifying key risk factors such as endoscopic ultrasound features, tumor margin clarity, diameter, and monocyte-to-lymphocyte ratio. Tested with 318 patients, the logistic classification model achieved high performance with area under the curve values of 0.919 and 0.925 for training and test sets, respectively. This tool supports personalized decision-making and enhances preoperative risk assessment in clinical settings.