MLiNS Laboratory
MLiNS (Machine Learning in Neurosurgery) Laboratory is a University of Michigan research laboratory led by Todd Hollon. It develops machine intelligence for human health and disease, especially neurological and neurosurgical conditions, translating advances in AI, machine learning and computer science into improved diagnosis and treatment.
At a Glance
- Academic and clinical research institutions
- Neurosurgery and neuro-oncology
- Radiology and neuroradiology
- Hospitals and health systems
- +2 more
AI Tools by MLiNS Laboratory
(1)
NeuroVFM
Neuroimaging Foundation Model MRI CT
Discussions
No discussions yet
Be the first to start a discussion about MLiNS Laboratory
Latest News
Optical microscopy predictions of focal recurrence in glioblastoma, a collaboration with the Hervey-Jumper Lab at UCSF using FastGlioma, published in Science Advances.
MLiNS announced the addition of students Annie Huang and Prateek Pinchi.
Akhil Kondepudi's research briefing on learning from routine health-system data to build better neuroimaging AI models appeared in Nature Medicine.
Todd Hollon's editorial, Benchmarking the Brain's Blood Vessels, was published in NEJM AI.
Products & Services
A generalist volumetric foundation model for multimodal neuroimaging. It uses Vol-JEPA self-supervision on 5.24 million clinical MRI and CT volumes and provides a 3D vision-transformer encoder, study-level diagnostic heads, findings generation and optional reasoning/triage workflows. Research use only; model weights are distributed through Hugging Face with access approval.
A visual foundation model for brain MRI trained at health-system scale for diagnosis, triage and clinically grounded decision support.
A computational pathology/visual foundation model for rapid, label-free detection and scoring of glioma infiltration at the surgical margin using stimulated Raman histology.
An AI system that predicts the molecular genetics of diffuse gliomas from rapid, label-free optical imaging within minutes of biopsy in the operating room.
Market Position
MLiNS differentiates itself by combining clinical neurosurgery with health-system-scale learning from routine MRI/CT and label-free intraoperative optical microscopy. In neuroimaging, its models are positioned against alternatives such as HLIP, PRIMA, NeuroMAE, DINOv3 and BiomedCLIP; in intraoperative pathology, FastGlioma and DeepGlioma target rapid, clinically grounded tumor assessment where conventional pathology workflows are slower.
Leadership
Founders
Todd C. Hollon
Neurosurgeon-scientist, Associate Professor of Neurosurgery at the University of Michigan, Joseph R. Novello and Alfredo Quiñones-Hinojosa Research Professor of Neurosurgery, and Program Director for Artificial Intelligence in Neurosurgery. He is the MLiNS principal investigator and leads research in machine intelligence, biomedical imaging and neurosurgery.
Executive Team
Todd C. Hollon
Principal Investigator; Associate Professor of Neurosurgery; Program Director, Artificial Intelligence in Neurosurgery
University of Michigan neurosurgeon-scientist and research professor. He leads MLiNS research spanning AI, biomedical imaging, machine learning and clinical neurosurgery.
Founding Story
The lab was formed around Todd Hollon's vision of bringing modern AI, machine learning and computer vision into clinical neurosurgery and biomedical research. Its initial and continuing aim is to develop clinically grounded systems that improve diagnosis and treatment of neurological disease, including intraoperative brain-tumor diagnosis and neuroimaging decision support.
Business Model
Revenue Model
MLiNS is an academic research laboratory rather than a disclosed commercial company. Its public offerings are research software, open-source repositories, interactive demonstrations and research models; no commercial subscription or usage-based pricing model was identified.
Target Markets
- Academic and clinical research institutions
- Neurosurgery and neuro-oncology
- Radiology and neuroradiology
- Hospitals and health systems
- Biomedical imaging and computational pathology researchers
- Medical AI and foundation-model developers
- Neuroimaging diagnosis and triage
- Radiology report and findings generation
- Clinical decision support for brain and spine disease
- Intraoperative brain-tumor detection and surgical-margin assessment
- Molecular classification of diffuse gliomas
- Tumor recurrence and patient-outcome forecasting
- Michigan Medicine
- University of Michigan clinical and research users