Online neuron reconstruction

Abstract

In this thesis, I designed an algorithm that traces neurons through the images of various datasets extracted from mouse brains. The algorithm is implemented in Python and relies on the output of a recently developed Fully Convolutional Neural Network (implemented in C) that runs on the underlying images. The images themselves are generated using state of the art technologies that provide high resolution and extremely accurate representations of the original brain matter. The algorithm is part of the combined efforts of MIT's Computational Connectomics Group.