The brain is one of the most complex and interesting systems available to human study. Modern technology allows for the brain to be studied in ever increasing detail, however analysis techniques still cannot determine effectively whether or not a given state is "healthy" or not. The proposed method is the use of a correlation matrix to extract a functionally connected network of brain regions from fMRI data, followed by the application of several generic methods of data analysis. This was executed on a dataset of 477 subjects with 216 ADHD patients and 261 normal control. Various thresholds for the correlation matrix as well as various parameters for the analysis methods were used. Overall the results, while not statistically viable, do indicate the utility of the methods proposed.
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.