Using Artificial Intelligence to Predict Gastric Cancer Risk in Adults with Intestinal Metaplasia or Atrophic Gastritis

What this Trial Is About

Researchers are studying patients with intestinal metaplasia IM and atrophic gastritis AG to identify clinical or tissue-based factors linked to the development of gastric cancer. The study aims to create a machine learning tool to predict future gastric cancer risk and help classify patients based on individual risk levels. This research is a two-part retrospective observational study. Clinical data from patients with histologically confirmed IM or AG will be collected from records dating between 2000 and 2020. Histology slides will also be reviewed using whole slide imaging to help develop and validate the machine learning prediction model. The study will include a training group with about 70% of cases and a separate validation group with the remaining cases. Participants medical history and tissue samples will be analyzed using computer algorithms to measure the occurrence of gastric cancer and dysplasia over 20 years. The study will assess the accuracy, sensitivity, specificity, and predictive values of the machine learning model. This long-term analysis will provide insights into gastric cancer risks in patients with IM and AG. The total study duration allows for comprehensive data collection and evaluation.

Prediction of Gastric Cancer in Intestinal Metaplasia and Atrophic Gastritis