Current research
Active Projects
Selected initiatives that demonstrate the lab's focus on auditable, explainable, and trustworthy biomedical AI.
Neuro-Symbolic AI for Drug Repurposing
This NIH-funded project develops hybrid neuro-symbolic AI methods to identify and evaluate potential therapeutic uses of existing drugs. It combines biomedical knowledge, clinical evidence, and machine learning to support more explainable and reliable drug-repurposing discoveries.
Explore portfolioNarrative-Driven AI for Clinician Burnout Surveillance
This project investigates how clinical narratives and healthcare data can reveal early signals associated with clinician burnout. The goal is to develop responsible and interpretable AI methods that support organizational awareness and timely intervention.
Explore portfolioTrustworthy Algorithms for Extreme Multi-Label Medical Coding
This project develops trustworthy and hierarchy-aware algorithms for assigning medical codes from large and complex coding systems. The work focuses on extreme multi-label classification, explainability, structural validity, and improved performance for rare clinical codes.
Explore portfolioA Socio-Technical Approach to Biomedical Content Authoring and Publishing
This project explores how researchers, domain experts, semantic technologies, and intelligent systems can work together to improve biomedical content authoring and publishing. The goal is to make scientific content more structured, interoperable, reusable, and accessible to both humans and machines.
Explore portfolio
