Position: AI-Driven Risk Stratification is Essential for Affordable Early Detection of Cancer

Asif Khan, Duncan Terell Forster, Moshir Harsh, Daniel Ritter, Chunlei Zheng, Rose Orenbuch, Abdullah Kuziez, Artem Gazizov, Debora Susan Marks, Nathanael R. Fillmore, Chris Sander

Keywords: Cancer risk, early detection of cancer, AI for medicine, GenAI for healthcare

Abstract: Early detection of cancer offers the best chance for cure, yet population-wide screening remains either impractical or insufficiently implemented for most malignancies due to several factors. These include heterogeneity of risk, limited technology for early detection, rarity of cases, lack of funding, and limited societal acceptance. This position paper argues that AI-driven risk stratification using longitudinal, population-scale electronic health records (EHRs) provides a practical first step toward scalable surveillance programs. We outline a three-stage framework of prediction, detection, and intervention, in which AI-based risk models nominate small high-risk cohorts to receive advanced diagnostic tests, benefit from emerging early detection technologies, and access timely therapy or preventive care. We emphasize the importance of evaluation metrics such as positive predictive value (PPV) and standardized incidence ratio (SIR), which reflect real-world feasibility, cost-effectiveness, and alignment with healthcare system capacity. As an example, we trained a pancreatic cancer risk model on the US Veterans Affairs database of 15.9 million patients, demonstrating that focusing on high-risk individuals can support a realistic surveillance program. We argue that AI-driven risk stratification, when deployed appropriately as decision support and integrated into early detection and intervention workflows, has the potential to transform cancer care through coordinated efforts across research and healthcare systems.

Submission Number: 90