Recent publications from the lab
October 20, 2026
It’s been a productive stretch for MARIA. A few highlights from our newest work:
- Synthetic gait video for privacy-preserving health AI. Doctoral researcher Tomi Giwamogorewa, with Julie Ries and Nathan Green, presented a protocol for generating synthetic gait videos to train gait-disorder models without exposing patient data, at the 2025 IEEE ICDM Workshops. This work is part of our NIH AIM-AHEAD project.
- Personalized falls prevention. Mahederemariam Dagne, Nathan Green, Diane Murphy, and Patricia Heyn analyzed outcomes from over 200,000 participants in evidence-based falls prevention programs and proposed a human-in-the-loop AI approach for recommending programs to individuals (Innovation in Aging, 2025).
- Research trends in cybersecurity doctorates. Diane Murphy, Nathan Green, and Donna Schaeffer used topic modeling on more than 100 D.Sc. dissertations to map what applied cybersecurity doctoral research looks like (IEEE FIE 2025).
- Machine learning for juvenile justice. Nathan Green and Erik Alda applied machine learning to optimize the surveys used to guide early intervention in juvenile justice (Empirical Economics Letters, 2025).
- LLMs and autocomplete in the metaverse. A new paper on using 3D space to integrate LLM suggestions and autocomplete into communication in virtual worlds appears in the CHIRA proceedings (Springer CCIS, 2026).
See the full list on our research page.