SUMMARY
Mayo Clinic researcher Shant Ayanian, M.D., uses advanced analytics and large-scale clinical data to improve risk stratification, treatment decision-making and care delivery. His research applies machine learning, clinical artificial intelligence (AI) and multisite data analysis to support better outcomes for adults with type 2 diabetes and for patients receiving hospital-based care.
Dr. Ayanian develops and evaluates predictive models, assesses their performance across diverse clinical settings and studies how data-driven tools can be integrated into routine clinical workflows. His studies address key challenges in model generalizability, data harmonization, implementation, governance, evidence synthesis and uncertainty assessment. In addition, he contributes to the development of foundational AI methods for genomic sequence analysis, extending his computational research beyond the electronic health record and clinical data.
Focus areas
- Diabetes predictive analytics. Dr. Ayanian studies electronic health record data from adults with type 2 diabetes to identify patterns associated with hyperglycemia, hypoglycemia and undertreatment. He uses machine learning and multisite clinical data to develop interpretable risk measures that may help healthcare professionals identify patients who could benefit from earlier, more personalized interventions.
- Clinical machine learning deployment. Dr. Ayanian investigates how predictive models perform when moved from development environments into routine patient care. His work examines generalizability across clinical sites, data drift, workflow fit and clinician usability to discover factors that influence whether machine learning tools remain effective and reliable in practice.
- AI governance. Dr. Ayanian explores the systems and processes needed to support the safe use of AI in healthcare. He examines model monitoring, oversight, maintenance and alignment with clinical workflows, emphasizing that successful implementation requires both strong technical performance and responsible integration into care delivery.
- Computational biomedical methods. Dr. Ayanian applies data science methods to areas beyond clinical prediction, including evidence synthesis, Bayesian analysis and genomic sequence modeling. He contributed to the development of STRAND, a foundational sequence transformer for nucleotide decoding. This reflects his broader interest in computational approaches that generate clinically and biologically meaningful insights from complex data.
Significance to patient care
Healthcare data can help clinicians identify patients who may be at risk of complications before those issues become more serious. Dr. Ayanian's research focuses on transforming electronic health record data into practical tools that support earlier detection, personalized treatment and closer monitoring when needed. He also studies how AI tools perform in real-world clinical settings to help ensure they remain accurate, reliable and safe. Together, these efforts may support more-informed care decisions, reduce preventable complications and improve patient outcomes.