Project Title: Deep Learning for Neural Structure Identification



Joel Welling, Sr. Scientific Specialist


We are exploring a possible approach to the identification of un-oriented structures in 3D data, with a focus on the neural network of the larval zebrafish.  
We have a large dataset of training and testing examples collected from Hildebrand et al.’s zebrafish, courtesy of Art Wetzel, one of their collaborators and a PSC scientific specialist.  In a series of previous internships, Richard Zhau, Minyue Fan, and Brian Leonard have tuned and tested a novel neural network that picks out the known neurons in this data. 
The goal of this project is to improve on that network, and if possible to extend the work to identify other 3D biological structures.  The network is based on spherical harmonic analysis of the input data.

Required Background

Strong Python programming skills; a willingness to deal with math.

Recommended Background

The student will ideally be familiar with orthogonal series expansion, an example being Fourier series.  



Learning Focus

The student will deepen their skills with Python and deep neural networks.  The student will learn some math, specifically relating to spherical harmonics.

Desired Results

The student will gain experience with deep learning, and learn methods for quantitative analysis of the results of numerical experiments.

Desired Major

The student would be a major in Computer Science, Neuroscience, Physics, or a related discipline.


The student in this position will receive an hourly wage.


To apply please submit your resume and cover letter to Vivian Benton,
The deadline to apply is March 31, 2020.