Triage in acute ischemic stroke involves several CT based imaging studies that provide information needed for deciding upon the optimal treatment for patient. First is a high resolution 3D CT scan of the brain without contrast. This is followed by a CT perfusion scan, which generates a sequence of 3D brain scans that shows the perfusion of contrast through the brain tissue. From these time series of 3D scans, one then computes three images: time-to-peak, blood volume, and blood flow. These images are then used to determine the extent of brain tissue that can be salvaged (penumbral) through thrombectomy (an endoscopic surgical procedure for removing blood clots in order to restore blood flow to the brain). Taking the perfusion CT scan is, however, time-demanding and time to intervention is critical in treatment of acute ischemic stroke. The main goal of this MSc thesis project is to use deep learning to predict the time-to-peak, blood volume, and blood flow images directly from the initial non-contrast CT scan using deep learning.