A Machine Learning-Based Survey of Cerebrovascular Disease Prediction

Abstract

Machine Learning (ML) is a branch of Artificial Intelligence (AI) that employs software implementations to examine the highest level of accuracy. ML can be applied to predict diseases in the health sector. When the blood flow to a portion of the brain is interrupted or diminished, brain tissue is deprived of oxygen and nutrients, resulting in a stroke. Within a minute, brain cells begin to die. There are two types of brain stroke: ischemic stroke (a blocked artery in the brain) and hemorrhagic stroke (a blood vessel leaks or bursts). The goal of this research is to implement and examine the ML algorithms that are employed in stroke prediction. This review represents the ML approaches utilized for stroke predictions, using previous studies. The death rate, morbidity, and functional result are all predicted outcomes, according to the majority of the studies. The most commonly used techniques to predict the stroke are Support Vector Machines, Random Forest, Decision Trees, Logistic Regression, KNN, XGBoost, and Artificial Neural Networks. Best results are produced based on the determination of precise attributes to utilize as causes of stroke. The purpose of this survey is to predict symptoms and changes in patient’s health at an early stage so that stroke can be observed later. For the prevention of major causes of stroke, the prime time of 0-90 minutes will be regarded as the prime period. Despite this, just a few oracles and classifiers produced reporting criteria for medical sector tools, none of which were useful. As a result, the goal of this analysis was to examine the accuracy of several ML algorithms for stroke prediction