We aim to develop fully automated algorithms for mapping networks within biological brains.

The field of connectomics aims to describe and characterize neurons in the brain, and the connections between them, by applying computer vision techniques to image data obtained from microscopes. Historically, extensive manual proof-reading has been required to curate the output of the machine learning pipeline. This represents an impractical bottleneck for a problem involving massive amounts of data. In our work, we seek to automate human intuition to develop a more thoroughly scalable approach to characterizing neural circuits.