Encephalic Stroke Prediction Using Machine Learning Algorithm

Detecting brain strokes is challenging due to complex visual patterns and subtle variations in affected regions. Traditional machine learning models, like Support Vector Machines and Decision Trees, struggle to capture these features accurately. Stroke prediction has advanced from clinical tools like the Hachinski Ischemic Score to AI-driven models, enhancing diagnostic precision. Recent deep learning and ensemble methods achieve up to 94% accuracy. Integrating Convolutional Neural Networks (CNNs) with XGBoost automates feature extraction, while Spiking Neural Networks (SNNs) offer superior performance with 95.6% accuracy. Inspired by biological systems, SNNs improve efficiency and stroke pattern recognition. Contour-based severity analysis further refines predictions. Combining CNNs for deep learning, XGBoost for classification, and SNNs for precision enhances both accuracy and reliability, contributing to early diagnosis and severity assessment.