Predicting Possible Skeletons for 3D Binary Segmentations of Neurons (Jun 2019 – Present)

I am working on a project to develop supervised machine learning algorithms that predict possible 3D skeletons for segmentations of electron microscopy images.

Drawing the skeleton of neurons is fundamental and vital to understanding the shape of neurons from a mathematical and statistical perspective. However, machine learning algorithms for predicting skeletons in 3D volumes have yet to be created. This project aims to create machine learning algorithms that can predict possible 3D skeletons of electron microscopy images of neurons. We hope that these algorithms can improve upon currently used skeletonization techniques.