Machine learning analysis of whole mouse brain vasculature

Journal: Nature Methods

Published: 2020-03-11

DOI: 10.1038/s41592-020-0792-1

Affiliations: 9

Authors: 12

Go to article

Research Highlight

AI offers window into brain vasculature

© Callista Images/Getty

© Callista Images/Getty

A technique that pairs artificial intelligence with methods for making tissues transparent could help scientists better understand how abnormalities in the brain’s blood flow contribute to stroke, dementia and other neurological conditions.

Changes to the structure of the blood vessels in the brain are indicative of many neurological disorders, including Alzheimer’s disease, stroke and brain injuries.

Now, a team co-led by researchers at the Technical University of Munich has used special dyes and chemical stains to make mouse brains see through. They then applied high-resolution imaging, followed by deep-learning approaches, to analyse all the brain’s blood vessels down to the micrometre scale.

In this way, they created a reference map of brain vasculature in three standard lab strains of mice, which all showed slight differences in their vessel architectures.

Supported content

  1. Nature Methods 17, 442–449 (2020). doi: 10.1038/s41592-020-0792-1
Institutions Share
TUM Department of Informatics, Germany 0.33
Institute for Stroke and Dementia Research (ISD), LMU, Germany 0.21
TUM Center for Translational Cancer Research (TranslaTUM), Germany 0.13
TUM Cluster of Excellence - Origin and Structure of the Universe (Excellence Cluster Universe), Germany 0.10
TUM Munich School of Bioengineering (MSB), Germany 0.08
Institute for Tissue Engineering and Regenerative Medicine (iTERM), HMGU, Germany 0.06
Institute of Pharmacology and Toxicology, UZH, Switzerland 0.04
Graduate School of Systemic Neurosciences (GSN), LMU, Germany 0.03
DZIF Partner Site Munich, Germany 0.03