Enhancing Stroke Risk Prediction: Leveraging Machine Learning and Magnetic Resonance Imaging Data for Advanced Assessment

As a leading cause of death, strokes have been regarded as a dangerously impactful condition with little to no predictability. Currently, there is no effective method to predict a stroke using warning signs and hereditary factors. We developed a quantitative method to predict strokes before happening. We used MRI scan data obtained from OpenNeuro, specifically images showing the signs of pre-stroke and post-stroke. We trained machine learning models using the data, including support vector machines (SVMs), K-nearest neighbors (KNNs), and random forests. The models predicted the risk of a stroke accurately. The models allow for diagnosing and enable clinicians to care for patients promptly, potentially saving lives and improving outcomes.