The aim of this study is to investigate the feasibility of using dual-energy (DE) imaging for discrimination of gray-white matter with brain CT. Initially, an optimization study was accomplished that aimed to find the optimal low and high beam energy spectra for the DE application. Simulations were carried out with 181 noise free brain dual-energy CT (DECT) images, simulated with an in-house developed software simulator for several monochromatic and polychromatic spectra within a full gantry acquisition arc of 360
. Ten monochromatic beams with energy from 20 keV to 110 keV and four polychromatic beam spectra 80, 100, 100, 140 kVp were simulated. The software brain phantom was the Digital Brain Phantom II, that is voxel-based model of the brain with no tumor insertion. To obtain DE projections, a non-linear algorithm for combining low and high energy images was exploited. Further, the optimal parameters of this algorithm were determined. The optimization study showed that the maximum contrast in brain DECT is achieved for the low/high energy combination of 100/110 keV in case of monochromatic beams and 80/100 kVp for polychromatic beams. For these cases, the contrast improvement was of the order of 12 and 7 times fold, respectively, compared to brain CT images acquired at high energies. The visual observation demonstrated superiority of the brain DECT compared to single energy brain CT images in terms of improvement in brain tissue differentiations. Evaluation of line profiles taken through different regions of interest on single and brain DECT images showed an excellent gray and white matter tissue discrimination.