Separation of veins from activated brain tissue in functional magnetic resonance images at 1.5 T

A feasibility study was conducted to segment 1.5 T functional magnetic resonance images (fMRIs) into grey matter and large veins using individual pixel intensity difference and temporal phase delay as two correlated parameters in 1.5 T gradient echo images. The time-course of each pixel in gradient echo images acquired during visual stimulation with a checkerboard flashing at 8 Hz was correlated to the stimulation `on´-`off´ sequence to identify activated pixels, and the temporal delay of each activated pixel was computed by fitting its time-course to a reference sine function. A histogram of the product pixel-intensity x temporal delay could be fitted to a bimodal distribution, which was then used to segment the functional image into veins or activated brain tissue. The results show relatively good demarcation between large veins and activated grey matter using this method