A Survey on Deep Learning Techniques for the Stroke Prediction on Neuro Images

Stroke is still one of the leading causes of death and chronic disability worldwide, and so immediate and proper diagnosis is a must to prevent delays in clinical management. Manual interpretation of imaging by conventional diagnostic methods is, by and large, time-consuming and susceptible to human bias. The medical imaging process has undergone a transformation in contemporary times due to the confluence of AI, ML, and DL approaches. Through computer-aided processing of CT and MRI scans, these have been shown to offer enormous potential to improve the detection, classification, and prediction of strokes.This survey paper provides a state-of-the-art overview of the AI-based models reported during the period from 2021 to 2024 in their methodologies, datasets, performance metrics, and clinical usability. By combining these contributions, the paper concludes that there are research gaps identified in current literature, including generalizability of models, a lack of standard datasets, and explainability requirements in clinical applications. This study aims to provide a foundation for further research into creating clinically integrated, interpretable, and reliable AI systems for stroke detection. Index Terms- Stroke Detection, Deep Learning, Machine Learning, Neuroimaging, CT Scans, MRI, Artificial Intelligence, Medical Diagnosis, Hemorrhagic Stroke, Ischemic Stroke, Healthcare AI, Feature Extraction, Image Classification.