Roy Gabriel, PhD Candidate
Medical Imaging AI • Multimodal Fusion • Cardiovascular Risk Prediction
About Me
I am a PhD candidate in Electrical and Computer Engineering at Georgia Tech, advised by Dr. Ali Adibi. My research focuses on medical imaging AI, clinically grounded representation learning, multimodal fusion, and long-horizon cardiovascular risk prediction. I develop AI systems that learn from complex clinical and imaging data, with recent work spanning CT-based prediction of major adverse cardiac events, probability-level multimodal fusion, active learning, cross-modal prediction of coronary artery calcium from chest X-rays, human-in-the-loop eye-tracking systems, and emerging work in women's health and fairness in medical AI. Outside the lab, I mentor students, build educational programs through Makademy, play piano, and enjoy soccer.
Research Snapshot
CARDINAL
CT Representation Learning for Cardiac Risk
- Clinically grounded representation learning from non-contrast CT
- Predicts 10-year major adverse cardiac events beyond traditional biomarkers
- Improves long-horizon risk prediction over standard clinical and imaging baselines
Multimodal Cardiovascular AI
Imaging + Clinical Data
- Multimodal framework achieving AUROC up to 0.88 on rare cardiac outcomes
- ~30% improvement over clinical and imaging baselines
- Probability-level fusion for robust, interpretable rare-outcome prediction
Data-Efficient and Cross-Modal AI
Label Efficiency and Accessibility
- Active learning reduced labeling burden by over 75%
- Chest X-ray to coronary artery calcium prediction up to 80% AUROC
- Research in women's health and fairness in medical AI
Experience Snapshot
Founder & Owner
Makademy LLC (May 2021 – Jan 2026)
- Built an ML-in-Finance course for IvyLine, reaching 90+ universities and 10 countries
- Developed an automated derivatives trading algorithm
- Scaled operations with a 95% average profit margin
Quant Research Intern
Dimensional Fund Advisors (Jun – Aug 2023)
- Automated ETF basket analysis on a 1.5 TB dataset
- Reduced storage requirements by 85%
- Delivered portfolio and basket management recommendations
FX Trading Intern
BNY Mellon (Jun – Aug 2022)
- Priced FX spot, swaps, and options
- Backtested volatility strategies in Python
- Built a Bloomberg API trading book for future interns