Connectomics is the study of maps and connections in an organism’s nervous system. The complexity and large variety of cells in the nervous system makes its comprehensive mapping infeasible by manual tracing of cells. We think that it is possible to approach this problem computationally. There are number of algorithms to segment brain images into neuronal processes. While these algorithms identify the large objects in most cases, they often leave out small parts of cells such as dendritic spines, which may play a significant role in determining the correct connections between cells. The goal of my project is to improve the quality of brain image segmentation by specifically finding and correcting mislabeled objects representing dendritic spines.

I am excited to be part of a project that tries to make a step towards understanding the brain, and in the long run can have a huge impact in medicine. I have done internships at Google and D.E. Shaw. Also started working on this project in summer and now I have a better idea of how to choose classes that will fit my project best.