Automatic segmentation of infarcted regions from Computed Tomography Perfusion Imaging using 3D Convolutional Neural Network

Luca Tomasetti

Abstract

A cerebral stroke is the second most common cause of death among adults worldwide: more than 13 million people suffer from this type of disease each year. Time is a fundamental factor: the treatment window for thrombolysis is estimated on 4.5 hours from symptom onset. One of the most used technique to deal with patients affected by cerebral stroke is Computed Tomography Perfusion (CTP) imaging: a fast method that allows to accurately predict the prognosis of patients in the early stage of the treatment.