Development of a Novel Deep Convolutional Neural Network Model for Early Detection of Brain Stroke Using CT Scan Images

Abstract

The importance of early brain stroke detection cannot be overstated in terms of patient outcomes and mortality rates. Although computed tomography (CT) scan images are frequently used to identify brain strokes, radiologists may not always be accurate in their assessments. Since the advent of deep convolutional neural network (DCNN) models, automated brain stroke detection from CT scan images has advanced significantly. It's probable that current deep convolutional neural network (DCNN) models aren't the best for detecting strokes early on. The authors present a novel deep convolutional neural network model for computed tomography (CT) images-based brain stroke early detection. The ability to extract features, fuse those features, and then recognize strokes is key to the proposed deep convolutional neural network model. To extract high-level information from CT scan images, a feature extractor with numerous convolutional and pooling layers is used.