Employing Machine Learning in A Multimodal Approach for Brain Stroke Modelling and Identification

All living things need to be healthy. Any disease can be avoided, which is always preferable. Heart attacks and strokes are two diseases that can be avoided. Most strokes are caused by an unexpected obstruction of the heart and brain channels. One of the deadliest disorders, brain stroke is now a frequent occurrence. A stroke may result in death, serious disability, or permanent brain damage. A stroke necessitates urgent medical attention. Early detection of many stroke warning signals can reduce the severity of a stroke. As a result, a lower death rate is possible. The data gathered from the subjects/patients in this study is split into two categories: dependent variables and non-dependent variables. These dependent and non-dependent variables are clustered and applied with the different ML algorithms such as Support Vector Machine (SVM), Gaussian Naive Bayes (GNB), Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), Light GBM (LGBM) and XG Boost (XGB) to predict the results. The stroke can be predicted by the suggested method and classify the type of the Ischemia stroke such as thrombotic stroke or embolic stroke. The model is able to predict possibility of occurrences of recurrence stroke by giving the warning message.