Important clinical signals can disappear inside electronic medical records. Rare diagnoses, uncommon patterns, and subtle combinations of codes are often diluted or lost when complex data is simplified for analysis. For certain patients—especially younger people who have never smoked—this can mean delayed recognition of elevated lung cancer risk.

A new collaboration between the OSF HealthCare Cancer Institute and Illinois State University is using artificial intelligence to address that challenge.

The project is supported by the Connected Communities Initiative (CCI), a strategic agreement between OSF HealthCare and Illinois State University that brings together the talents of both organizations to tackle real-world health care challenges in our communities. Through multidisciplinary teams of clinicians, faculty, and researchers, CCI focuses on areas such as data science, clinical education, health care engineering, and cybersecurity—with the goal of developing practical solutions that can improve how care is delivered.

This particular project is testing whether large language models can help preserve clinically meaningful information that traditional approaches frequently overlook. It is co-led by Jun Zhang, MD, PhD, Vice President of Oncology Research at the OSF HealthCare Cancer Institute; Jonathan Handler, MD, Senior Fellow of Innovation Clinical Intelligence at the JUMP Simulation Center; and Meenal Chaudhari, PhD, Assistant Professor in the College of Applied Science and Technology at Illinois State University.

The team is comparing different methods of organizing the complex and often rare clinical codes found in medical records—expert-driven approaches, statistical techniques, and AI-informed methods that draw on large language models. Their initial focus is improving predictive models for lung cancer risk among younger nonsmokers, a group for whom standard risk tools are often less effective.

The goal is practical and patient-centered: develop more reliable tools that support earlier identification of individuals who may benefit from closer monitoring or further evaluation, while retaining the rare but important clinical details that matter in real-world care.

By combining deep clinical expertise with advanced computational approaches, OSF and Illinois State are working to turn complex data into clearer insight—ultimately helping deliver more precise and timely care to the communities we serve.

Read More about the Investigators

Jun Zhang, MD, PhD, Vice President of Oncology Research at the OSF HealthCare Cancer Institute

Jonathan Handler, MD, Senior Fellow of Innovation Clinical Intelligence at the JUMP Simulation Center

Meenal Chaudhari, PhD, Assistant Professor in the College of Applied Science and Tech