EEG-Driven Machine Learning for Stroke Detection in High-Risk Patients

Stroke remains a leading cause of disability and mortality worldwide, highlighting the need for effective tools for early detection and intervention. Recent research has explored the use of bio-signals generated by the human body as indicators of stroke occurrence. Among these, Electroencephalography (EEG) has shown particular promise. EEG-based stroke detection offers a non-invasive, cost-effective, accurate, and portable solution. This paper investigates the use of Machine Learning (ML) techniques with EEG data to detect strokes. Four experimental setups were designed to evaluate different feature engineering methods: using all features, selecting features via a Decision Tree (DT) with varying thresholds, and applying Principal Component Analysis (PCA) and Independent Component Analysis (ICA) for dimensionality reduction. Results indicate that the fourth setup—applying ICA with both AdaBoost and XGBoost—yielded the best performance, achieving an accuracy of 89%, precision of 86%, recall of 100%, F1-score of 92%, and a Matthews Correlation Coefficient (MCC) value of 0.76.