Cerebral Stroke Prediction Using Machine Learning Algorithms

Cerebral stroke is on the rise, which may kill, disable, and destroy the brain. In this situation, it is important to predict a cerebral stroke early to prevent or lessen the damage caused by a stroke. A cerebral stroke occurs when brain tissue is deprived of oxygen and nutrients due to decreased or blocked blood flow. Currently, machine-learning-based systems are widely used as an effective method for predicting and reducing the potential damage of various diseases. The goal of this research is to find the early signs of a cerebral stroke so that people can take steps to stop more damage. Here, a dataset is used with 5026 points of data, 11 features about stroke, and the five best machine learning models trained for making predictions: decision tree, random forest, KNN, XGBoost, and a neural network model. Compared to the other machine learning models, the random forest and XGBoost models performed better. The accuracy of Random Forest was 97.11%, whereas that of XGBoost was 97%. Thus, the most accurate model, Random Forest, is used to forecast the chance of a stroke. A hosted web application and a mobile app are created to make the system accessible. By facilitating early prediction and intervention, this study can improve the medical system’s capacity to prevent the damage of cerebral stroke.