Brain Stroke Prediction Using Machine Learning

DOI

https://www.doi.org/10.59256/ijrtmr.20260602053

Published In

Volume 6, Issue 2 (2026)

Published Online

March-April 2026

Pages

380-390

Article Type

Original Article

Check for updates

Brain stroke prediction is a critical concern in the healthcare sector due to its severe impact on human life, including mortality and long-term disability. Traditional diagnostic methods rely heavily on clinical expertise and manual analysis, which can be time-consuming and may not always support early detection. With the advancement of data-driven technologies, there is an increasing need for intelligent systems that can automatically predict stroke risk using computational techniques. This research focuses on developing a machine learning-based framework for accurate brain stroke prediction. The study utilizes a dataset containing patient health attributes such as age, hypertension, heart disease, average glucose level, body mass index (BMI), smoking status, and other demographic features. These attributes play a significant role in identifying individuals at high risk of stroke. Multiple machine learning models, including Logistic Regression, Support Vector Machine (SVM), K-Nearest Neighbours (KNN), Decision Tree, Random Forest, and XGBoost, are implemented and compared. Advanced preprocessing techniques such as data cleaning, handling missing values, normalization, and feature scaling are applied to enhance model performance. The models are evaluated using performance metrics including accuracy, precision, recall, and F1-score. Experimental results show that Logistic Regression achieved an accuracy of 94.98%, Decision Tree 91.47%, KNN 94.48%, SVM 94.98%, Random Forest 94.98%, and XGBoost 94.98%, where XGBoost provided the highest performance among all models. Furthermore, Explainable Artificial Intelligence (XAI) is incorporated using SHAP (Shapley Additive Explanations) to improve model interpretability. The analysis reveals that factors such as age, glucose level, hypertension, and smoking status have a significant influence on stroke prediction. This enhances transparency and builds trust in the model for healthcare applications. Over all, the proposed approach provides an efficient and scalable solution for early stroke prediction. It has strong potential for real-world applications such as clinical decision support systems, preventive healthcare, and risk assessment, thereby helping in reducing stroke-related complications and improving patient outcome.