Using AI with Electronic Health Records to Detect and Predict Cancer Early

What this Trial Is About

This research aims to evaluate an AI-assisted predictive model designed to identify and diagnose cancer by using a wide range of health data. It focuses on early cancer detection to improve patient outcomes and survival rates, integrating medical history, lab results, imaging scans, and genetic markers. The study builds on a previous phase that analyzed data from around one million cancer patients to develop initial models. The current phase is a prospective, multi-center observational study launched in February 2025. It involves two groups a healthy cohort without diagnosed cancer serving as a control group, and a tumor cohort of individuals diagnosed with various cancers. The AI system uses deep learning to analyze multimodal data from electronic health records, lab tests, imaging, and genetics to predict cancer risk and improve diagnostic accuracy. Participants will provide comprehensive health records including medical history and test results. Researchers will measure outcomes such as diagnostic accuracy using Area Under the Curve AUC and F1 Score over one year. Secondary outcomes include sensitivity and specificity of cancer detection. The study does not involve treatment but observes and compares data from both groups to validate the AI models effectiveness in real-world clinical settings.

AI-Driven Cancer Diagnosis and Prediction With EHR