Todd Hollon
MLiNS Research Programs
We develop machine intelligence that understands human health and disease, especially related to the nervous system. Our work is organized into five research programs, each with its own models, datasets, and publications.
AI-based Neuroimaging
Clinically grounded AI for neuroimaging learned from routine health system data at scale, including NeuroVFM, Prima, HLIP, and ItemizedCLIP.
Explore AI-based Neuroimaging →
Intelligent Histology
Stimulated Raman histology (SRH) combined with deep learning for rapid intraoperative brain tumor diagnosis, computational pathology, and tumor margin delineation.
Explore Intelligent Histology →
Visual Intelligence
Visual representation learning, visual reasoning, and vision-language modeling research, including HiDisc, CLIPred, SimCLIP, and CodeV.
Patient Forecasting
Interpretable machine learning for predicting surgical recovery, long-term outcomes, and survival from electronic medical record data.
Collaborative Neuro-Oncology
Team science in brain tumor research: shared biobanks, multi-disciplinary trials, and privacy-preserving collaboration such as federated learning.
Explore Collaborative Neuro-Oncology →
Interested in joining one of these programs? We are hiring PhD students and postdocs in medical AI. Please email tocho [at] umich.edu.