Boston University · AI & human health
Multimodal AI for neurodegenerative disease and human health
We are computer scientists, neuroscientists, engineers and physicians building methods that fit the science, not science that fits the methods. Our models diagnose dementia, estimate disease biomarkers and interpret medical images, and we release them openly.
What we work on
Four connected themes, one goal: AI that clinicians and scientists can trust and use.
Dementia diagnosis & neuroimaging
Multimodal models that bring together MRI, clinical and neuropsychological data to assess cognitive status and tell dementia etiologies apart.
Learn moreComputational pathology
Graph- and transformer-based learning on whole slide images, including fusion with gene expression to predict outcomes.
Learn moreDigital & voice biomarkers
Detecting cognitive impairment from voice recordings and digital cognitive assessments collected at scale.
Learn moreMedical language models & agents
Large language models and agentic systems for medical question answering, research and education.
Learn moreFeatured papers
Recent highlights. Code for most of our papers is on GitHub.
FourierMIL: Fourier filtering-based multiple instance learning for whole slide image analysis
AI-driven fusion of multimodal data for Alzheimer’s disease biomarker assessment
Active feature acquisition via explainability-driven ranking
AI-based differential diagnosis of dementia etiologies on multimodal data
Latest news
All news- May 2026 Milestone Chonghua Xue successfully defends his PhD in Electrical and Computer Engineering.
- March 2026 Milestone Diala Lteif successfully defends her PhD in Computer Science.
- March 2026 Milestone Olivia Zhou got matched to Orthopedic Surgery at Stony Brook Medicine.
- March 2026 Grant We received an R03 grant from the National Institute on Deafness and Other Communication Disorders.
- February 2026 Paper Agentic AI for medical question-answering tasks, published in the International Journal of Medical Informatics.
- December 2025 Paper Fourier filtering-based multiple instance learning, published in the International Journal of Computer Vision.
- December 2025 Paper AI-augmented frameworks for AD clinical trials, published in the Journal of Prevention of Alzheimer's Disease.
Build the next generation of medical AI with us
We form small teams with complementary expertise in computer science, mathematics, neuroscience and medicine, and work persistently on comprehensive solutions. PhD, MD and MD-PhD trainees, and postdocs, are welcome.