Optimizing Neuron Segmentation for Connectomics
Connectomics” refers to the study of connections in the human brain. High-resolution EM images of brain tissue allow us to reconstruct a 3-D map of the brain (neurons synapses) which can lead to greater understanding of brain physiology and disease. However the volume of images is on the scale of petabytes. My project aims to increase the performance of “segmentation”° “° S the multi-step process to extract neurons from images “° S with novel algorithms and their high-throughput implementation on state-of-the-art hardware.”

Research gets me out of the bubble of computer science. It’s refreshing to see an interdisciplinary approach to solving problems in our world and to challenge myself to contribute alongside phenomenal faculty at MIT.