LAUSR

Machine learning has been a topic of several scientific conferences in 2017 and its potential to aid the (neuro)radiologist is now widely accepted. The task is now to find applications so that our patients can benefit from machine learning. One possible application is the prediction of delayed cerebral ischemia (DCI) following acute aneurysmal subarachnoid hemorrhage. The low sensitivity of the Fisher scale in Fisher grade III patients and the superiority of the Hijdra scale has just been addressed in the September issue of Clinical Neuroradiology [1], when abstracts indicating the potential to predict DCI from dynamic susceptibility contrast (DSC) perfusion, digital subtraction angiography or computed tomography (CT) were presented at the 52nd Annual Conference of the German Society of Neuroradiology [2–4]. Vasospasm following aneurysmal subarachnoid hemorrhage is also a matter addressed at the conference [5–7]. New treatment strategies are needed and have to be carefully evaluated [7, 8]. Just one example indicating the potential of machine learning in clinical neuroradiology.