Daniel B. Rowe1, Luis Hernandez-Garcia2
1Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI, USA; 2FMRI Laboratory, University of Michigan, Ann Arbor, MI, USA
A computationally fast high tSNR magnitude and phase activation model is presented then applied to an ASL fMRI visual experiment. In fMRI, Fourier encoded k-space measurements, inverse Fourier reconstructed images, and voxel time series are complex-valued (real and imaginary or magnitude and phase). Nearly all fMRI studies derive functional activation from magnitude-only time series while discarding the phase time series. A GLM magnitude and/or phase activation model has been introduced and shown to have higher sensitivity. However, the existing method to compute magnitude and phase activation utilizes an iterative MLE algorithm while here we use a large tSNR exact solution.