Washington University in St. Louis has established a research institute focused on using artificial intelligence to address critical health issues. The AI for Health Institute, directed by Chenyang Lu, brings together faculty from across the university.
Lu, the Fullgraf Professor of Computer Science & Engineering at Washington University in St. Louis, with joint appointments in Anesthesiology and Medicine, explained that the institute was formed three years ago to connect AI researchers with health professionals. A Fellow of the ACM and IEEE, Lu received the 2022 Outstanding Technical Achievement and Leadership Award from the IEEE Technical Community on Real-Time Systems and serves as the editor-in-chief of ACM Transactions on Cyber-Physical Systems.
Related: AI Enhances Healthcare with Deeper Insights
Research at the institute focuses on applying AI to solve important health problems using data-driven approaches. One initiative uses longitudinal electronic health records—including diagnostic codes and lab results—to predict conditions like cervical spondylotic myelopathy (CSM), a common spinal disorder, months before symptoms worsen. “Patients often endure prolonged suffering before receiving a diagnosis and treatment, which is more effective when administered earlier,” Lu noted.
To promote collaboration, the institute aims to eliminate barriers between AI researchers and health experts who typically work in isolation. “They lack familiarity, use different terminology, and have no history of working together,” Lu said. The institute supports teams in developing joint proposals and testing new concepts.
One major obstacle is maintaining AI effectiveness over time. Models trained on data from one hospital may underperform elsewhere, a problem Lu describes as the “spatial” challenge. Meanwhile, the “temporal” challenge involves models losing accuracy as patient populations or hospital practices change. “Physicians might notice declining accuracy, but this is a systemic issue,” he explained. “Monitoring model performance is essential to detect degradation.”
Related: Overcoming Remote Business Challenges With Advanced Techniques in 2026
Governance also plays a critical role. Many medical schools and hospitals now have AI committees to review models before deployment, ensuring they avoid bias and stay aligned with their intended purpose.
AI adoption in clinical settings has progressed more slowly than in administrative tasks like coding and revenue management. “Hospitals prioritize efficiency tools because they pose fewer safety risks,” Lu said. However, clinical decision support holds significant promise. Radiology has led the way, with AI tools flagging suspicious areas in imaging, while other specialties are integrating AI into electronic health record systems.
Medical education is evolving to include AI training. The university’s School of Medicine is incorporating AI education into its curriculum to prepare future doctors to assess AI outputs critically. “Physicians should leverage AI for efficiency while remaining alert to potential errors,” Lu advised.
Related: AI listening aids rural New Mexico hospital care
Regulation presents a complex challenge. Excessive rules could hinder progress, while weak oversight risks patient safety. “Both over-regulation and under-regulation carry risks,” Lu observed.
The institute’s approach reflects a growing trend in AI for health, where success relies on cross-disciplinary teamwork.
