CT perfusion-based prediction of final tissue status in cerebral ischemia

Abstract

CT Perfusion offers us some great diagnostic features for neuroimaging and has great potential in the field of acute ischemic strokes. Over the last few years more research is being conducted regarding the prediction of final tissue status in cerebral ischemia. The goal of this research is to automatically or semi-automatically differentiate between infarct core lesion and ischemic penumbra. Some commercially available packages already provide such features, but are not always as accurate. Much research is ongoing in more specialized institutes featuring different methods like thresholding, functional clustering and more advanced methods such as the use of artificial neural networks. Results look promising but we should remain critical before applying it in clinical settings due to the fact that most current research is being conducted on homogeneous data sets and is not clinically validated.