Estimation of T2* in functional spectroscopy during visual stimulation

Changes in T2* of activated pixels were estimated using single voxel and 1D CSI proton spectroscopy during visual stimulation. The single voxel was positioned to enclose activated pixels as determined by a corresponding functional MRI study. Angulated voxels represent the best match between the voxel boundaries and activated pixels. However, in the authors' 1.5 T clinical MRI, eddy currents in angulated voxel studies produced a bump in the water peak, which was removed by modeling a biexponential function with two frequencies to fit the time domain data. The results show approximately 1.8% T2* change when averaged over 5 subjects. Non-angulated voxels, which do not enclose the activated pixels as closely as the angulated voxels but allow estimation to be performed with a mono-exponential function gave similar T2* changes. Fitting also determined the amplitude of the water signal which is directly related to any T1 or in-flow related changes. The results show negligible changes in the amplitude. Also, peak height, which is directly related to T2*, was found to increase during stimulation by approximately 2% when averaged over 5 subjects. The peak height was also used to localize activation in the visual cortex in 1D CSI experiments. It is concluded that T2* changes reflecting a BOLD effect are demonstrated in these studies.