Research at AggieBrain

AggieBrain aims to create a publicly available repository of high-quality, expertly curated histopathology datasets that can be used by multiple research communities:

  • Computational health researchers can calibrate, stress-test, and improve their algorithms using standardized, openly available resources.
  • Neuropathology researchers can train domain-specific models that support more personalized disease assessment.
  • Global research teams gain access to tools and datasets that open the problem to the world, enabling new discoveries far beyond UC Davis.

As part of this effort, AggieBrain is also advancing new benchmarking methods to evaluate the reliability of emerging multimodal AI systems in pathology. These benchmarks are designed to identify and reduce common failure modes—such as factual or object-level inconsistencies—ensuring that future AI tools are both trustworthy and safe for high-stakes biomedical applications.

Together, these advances position AggieBrain not just as a data resource, but as a computational engine for discovery—shrinking the “haystack” of complex brain data so scientists can find the “needles” that matter most. By uniting neuropathology, engineering, and frontier AI, AggieBrain aims to accelerate progress toward better diagnosis, better treatments, and ultimately a dementia-free future.

Histology photo (left) with three colored segmentation maps (cyan, yellow, teal) with the grey matter/white matter, depicted in cyan and yellow respectively, detects plaques depicted in orange, and reports the counts in each region.
An example output of the AI-driven pathology quantification pipeline. Given a stained whole slide image of human brain tissue, the AI model automatically segments the grey matter/white matter, depicted in cyan and yellow respectively, detects plaques depicted in orange, and reports the counts in each region.

AggieBrain builds on years of successful collaboration between Brittany Dugger, Ph.D., Professor of Pathology and Laboratory Medicine and leader of the UC Davis Health Neuropathology Core, and Chen-Nee Chuah, Ph.D., Child Family Professor in the College of Engineering. Their prior work, supported by the Chan Zuckerberg Initiative DAF, an advised fund of the Silicon Valley Community Foundation, the National Institutes of Health, the National Science Foundation, the University of California Office of the President, the California Department of Public Health, and the Noyce Initiative, established AI-enabled pipelines for analyzing brain pathology. UC Davis is contributing to national, multi-institution efforts (such as the Brain Digital Slide Archive) to expand sharing of whole slide images alongside automated, quantitative approaches to understanding neuropathology and to support future precision medicine strategies. We are fortunate to work with a network of wonderful collaborators from Emory University, University of Kentucky, Human Computation Institute, UC San Francisco, and UC Irvine. You can read more about our current and past collaborative projects here.