fMRI scans yield important data about how brains respond to stimuli. The presence or absence of certain psychological disorders, such as autism, often affects these scans, making it possible for healthcare professionals to use fMRI data as a diagnosis or research tool. The signals read by fMRI, however, are mostly noise. Filtering of the important signals from the noisy signals, and then mapping codependent signals to each other has been done in the past using techniques such as statistical parametric mapping (SPM). Independent component analysis is used to isolate statistically independent signals. We have utilized ICA to create another filtering technique. Using this technique, we are able to find similar independent components within a patient's brain, and across multiple patients. To determine ICA's success in finding and mapping important signals, we have compared it to SPM data.
Advised by Linda Shapiro
CSE 203
Wednesday
May 21, 2008
3:30 - 4:20 pm