FMRI is a modern technique employed to help diagnose internal brain injuries. Analyzing the resulting data from them is currently facilitated by medical professionals. Having these professionals involved can increase the inherent cost of this process with regards to time and monetary value. For our project we aimed to address this problem by creating a neural network that can analyze unprocessed fMRI data and identify brain injuries. The result of this was a model that could predict whether a given fMRI contained a concussion with 70% accuracy.
Worcester Polytechnic Institute
Major Qualifying Project
All authors have granted to WPI a nonexclusive royalty-free license to distribute copies of the work, subject to other agreements. Copyright is held by the author or authors, with all rights reserved, unless otherwise noted.