DEEP LEARNING-BASED ISCHEMIC STROKE DETECTION USING CT PERFUSION IMAGING

Abstract

Stroke is one of the major causes of disability that represents a central public health concern. Early recognition and accurate diagnosis are pivotal for effective management of stroke clinically, as timely diagnosis significantly affects prognosis and reduces the risk of severe complications among patients. This research emphasizes ischemic stroke that is a particularly serious illness because it interrupts blood flow to a region of the brain, resulting in insufficient oxygen and nutrient supply to the brain. In this research work, deep learning methods are applied, and Computed Tomography Perfusion scans are used to detect ischemic stroke.  To analyze and evaluate the results, different deep learning methods like Convolutional Neural Networks, Recurrent Neural Networks, Artificial Neural Networks, and Deep Neural Networks are applied. The Convolutional Neural Networks exhibited significant results in the detection of ischemic stroke, with 91% accuracy, followed by Artificial Neural Networks with 71%