Contact
Email: yxw1259@miami.edu
Roles
Research Assistant Professor of Neurology
Secondary Faculty in Computer Science
-
Biography
Dr. Yusen Wu is a Research Assistant Professor in the Department of Neurology at the University of Miami Miller School of Medicine, with a secondary appointment as an Assistant Professor in the Department of Computer Science, where he teaches Data Security and Privacy. He was previously an Assistant Scientist at the University's Frost Institute for Data Science and Computing (IDSC). He received his Ph.D. degree in Computer Science from the University of Maryland, Baltimore County, where his dissertation focused on building scalable and trusted distributed systems.
Dr. Wu's research lies at the intersection of artificial intelligence, healthcare, and secure distributed systems. In the Department of Neurology, he develops multimodal models and large language models for the early diagnosis and personalized prediction of Parkinson's disease and other neurodegenerative disorders, drawing on neuroimaging, clinical records, and wearable and non-contact sensor data. Working in close collaboration with clinical practice, his goal is to advance precision neurology by building reliable AI frameworks that detect subtle early markers of disease onset and progression. Over the past several years, he has served as a lead investigator and developer on multiple federally funded projects for NASA, NSF, and the FDA, spanning AI modeling, drift detection, and blockchain applications. -
Education & Training
Education
-
Research Interests
Dr. Wu develops multimodal AI models for the early diagnosis and personalized prediction of Parkinson's disease and other neurodegenerative disorders, integrating neuroimaging, longitudinal electronic health records, clinical assessments, and wearable and non-contact sensor data. His work targets the prodromal window, where disease markers are subtle and distributed but intervention has the greatest potential value. A related line of research applies large language models to clinical text, extracting structured meaning from unstructured notes and generating interpretable explanations of model predictions, with emphasis on grounding and factual reliability.
Because models degrade once deployed, he also studies distribution and concept drift detection, model monitoring, robustness under population and site shift, and evaluation protocols that reflect real clinical deployment rather than static benchmarks. Building on his doctoral work on trusted distributed systems, he investigates privacy-preserving and verifiable computation for multi-institutional health data, including data provenance, integrity, and auditability, enabling collaborative model development across institutions that cannot pool raw patient data.
Across these threads, Dr. Wu works closely with clinical collaborators and treats deployability and clinician usability as research constraints rather than downstream engineering concerns. His long-term goal is clinical AI that is accurate, secure, interpretable, and trusted enough to change how neurodegenerative disease is detected and managed.