Interpretable AI for Cardiac Wellness & Equitable Healthcare

Building trustworthy AI systems grounded in axiomatically sound mathematical theory. My research spans cardiac stability theory, multimodal signal processing (ECG, PPG), explainable AI methods, physiologic invariance under drift, and longitudinal trajectory analysis. Applications include cuffless blood pressure estimation via attractor-vascular coupling theory and fairness-aware cardiac monitoring across diverse demographic groups.

Research Publications Lab

Research Focus

Cardiac Stability Theory & Longitudinal Monitoring

Axiomatically grounded framework defining health as dynamical stability on a bounded cardiac attractor. Extends via Physiologic Trajectory Theory (PTT) to define per-individual admissible corridors for non-invasive PPG monitoring and detection of meaningful cardiac drift.

Fairness & Demographic Equity in Signal Processing

Smartphone PPG processing across diverse skin tones (Fitzpatrick 1-6). IRB-submitted study collecting first public longitudinal PPG dataset. Addresses demographic bias in optical biomarkers through principled signal characterization, not post-hoc corrections.

Theoretical Foundations for Trustworthy AI

Nine interconnected theories (CST, CFD, EAT, PIT, PECT, CPGT, ADT, AVCT, PTT) grounding explainability, robustness, and physiologic validity. Each theory extends the last, creating a coherent ecosystem for high-stakes healthcare AI applications.

Research Overview

Research Vision

My research centers on a simple but profound question: What does it mean for a heart to be stable? The answer is Cardiac Stability Theory (CST)—a framework that defines cardiovascular health as maintenance of moderate dynamical complexity on a bounded attractor, measurable through Lyapunov exponents, recurrence structure, and entropy from any cardiac signal: ECG, PPG, or beyond.

From CST, I derive eight additional theories that address progressively harder questions: How do we preserve diagnostic information across signal domains (CFD)? When should models trust themselves (PIT)? How do we detect true drift from noise (PECT)? How do we synthesize data without losing structure (CPGT)? How does cardiac geometry predict blood pressure (AVCT)? And how do we track longitudinal wellness equitably across all demographic groups (PTT)?

This ecosystem—nine interconnected theories grounded in axiomatically sound mathematics—provides a principled, trustworthy foundation for high-stakes healthcare AI, where decisions must be transparent, robust to physiologic variation, and equitable across all human populations.

Teaching Philosophy

My teaching philosophy is grounded in the belief that active engagement leads to deeper learning. I design learning experiences that emphasize practical application, analytical reasoning, and hands-on problem solving. By connecting technical concepts to real organizational and industry contexts and helping students to develop both competence and confidence. I continuously work to enhance the learning environment through thoughtful curriculum improvement and evidence-based instructional practices. I also champion computer science for all and strive to ensure that every student has equitable access to high-quality computing education.