Deep Neural Network based Algorithm for Stroke Prediction

A Stroke is a medical condition where the brain's blood supply pathways burst, causing brain damage. When the brain's flow of blood and other supplies is hampered, adverse consequences may occur. According to the World Health Organization (WHO), stroke is the most common reason for mortality as well as disability. Rapid detection of the diverse warning symptoms of a stroke can assist in reducing the harsh-ness. This work presents a stroke predictive analysis procedure based on a deep learning model in which machine learning techniques are widely used. A convolutional neural network with an effective pre-processing step is employed in developing the model to accomplish the goal. Since the dataset contains several imbalanced data, the pre-processing model can effectively reduce considering the issue that improves the overall accuracy. According to experimental results, the proposed model achieves 96.8 percent accuracy, which is superior to the state-of-the-art model.