Dynamic perfusion imaging is routinely used in the diagnostic workup of acute ischemic stroke (AIS). At present, perfusion imaging can also be performed within the angio suite using flat detector computed tomography (FDCT). However, higher noise level, slower rotation speed and lower frame rate need to be considered in FDCT perfusion (FDCTP) data processing algorithms. The Time Separation Technique (TST) is a model-based perfusion data reconstruction method developed to solve these problems. In this contribution, we used TST and dimension reduction, where we approximate the time attenuation curves by a linear combination of trigonometric functions. Our goal was to show that TST with this data reduction does not impair clinical perfusion measurements. We performed a realistic simulation of FDCTP acquisition based on CT perfusion (CTP) data. Using these FDCTP data, we showed that TST provides better results than classical straightforward processing. Moreover we found that TST is robust to additional noise. Furthermore, we achieved a total processing time from reconstruction of FDCTP data to generation of perfusion maps of under 5 minutes. Perfusion maps created using TST with a trigonometric basis from FDCTP data show equivalent perfusion deficits as CT perfusion maps. Therefore, this technique can be considered a fast reliable tool for FDCTP imaging in AIS.