Scott Peltier1, Jonathan Lisinski2,
Douglas Noll, Stephen LaConte2
1Functional MRI Laboratory, University
of Michigan, Ann Arbor, MI, United States; 2Computational
Psychiatry Unit, Baylor College of Medicine, Houston, TX, United States
This
work examines support vector machine (SVM) classification of complex fMRI
data, both in the image domain and in the acquired k-space data. We achieve high classification accuracy
using image magnitude, image phase, and k-space magnitude data. Additionally, we maintain high
classification accuracy even when using only partial k-space data.