A New Approach for Brain Stroke Detection and Classification using Multi-layered Deep Neural Network

This study highlights the efficient procedure for detection and classification of Brain strokes with the help of Artificial Intelligence based medical image processing.Brain stroke detection and classification is still an open-ended problem and requires human intervention for accurate diagnosis.Leveraging a sophisticated neural network architecture inspired by AlexNet, the model demonstrates exceptional proficiency in identifying stroke cases and further classifying them into hemorrhagic and ischemic sub-types.The research involved comprehensive data acquisition and preprocessing, where medical images were meticulously categorized and normalized for training.Special attention was given to class weight balancing and data augmentation techniques to enhance generalization and prevent overfitting.The model achieved outstanding accuracy in both stroke detection and classification, showcasing over 94% accuracy in detection and 98% in classification.These results mark a significant advancement in medical imaging and AI applications in healthcare, offering a promising new tool for rapid and reliable stroke diagnosis and contributing to improved patient outcomes.