Artificial Intelligence Mortality Prediction Model for Gastric Cancer Surgery Based on Body Morphometry, Nutritional, and Surgical Information: Feasibility Study

The objective of this study is to develop a mortality prediction model for patients undergoing gastric cancer surgery based on body morphometry, nutritional, and surgical information. Using prospectively built registry from the Asan Medical Center (AMC), 621 patients, who were treated with no recurrence cancer, selected development model. Input features (i.e., surgical, clinicopathologic information) in collected data XGBoost analysis results experts’ opinions. A convolutional neural network (CNN) framework was developed predict surgery. Internal validation performed split datasets AMC, whereas external Ajou University Hospital. Fifteen survival probability suggestions. Accuracy, F1 score, area under curve our CNN 0.900, 0.909, 0.900 internal set 0.879, 0.882, 0.881 set, respectively. Our published website where anyone could using individual patients’ data. provides substantially good performance predicting mainly web application, clinicians will be able efficiently manage risk factors.