Using Deep Learning to Predict How Advanced Gastric Cancer Responds to Neoadjuvant Chemotherapy Based on CT Scans, Biopsy Images, and Clinical Information

What this Trial Is About

Researchers are developing a deep learning model to predict how well patients with advanced gastric cancer will respond to neoadjuvant chemotherapy. This model uses data from patients CT scans, biopsy pathology images, and clinical information to forecast the effectiveness of chemotherapy and the patients prognosis. The study aims to assist personalized treatment decisions for gastric cancer patients by improving prediction accuracy. Participants diagnosed with advanced gastric cancer will undergo standard neoadjuvant chemotherapy according to clinical guidelines. Data collected include CT imaging taken within one month before treatment, pathology images from gastric tumor biopsies, and detailed clinical profiles such as tumor markers and staging. The study includes both retrospective data from 1,800 patients used for model training and validation, and prospective data from 200 patients for model performance evaluation. During the study, researchers will gather clinical, imaging, and pathology data to develop and test the AI model. The primary outcomes are the accuracy and area under the curve AUC for predicting tumor regression grade TRG within two months. Secondary outcomes include progression-free survival at three years and overall survival at five years. Participants treatment details and prognosis will be closely monitored, with all study procedures designed to improve future personalized care for gastric cancer patients.

AI Prediction of Gastric Cancer Response to Neoadjuvant Chemotherapy