Predictive risk modeling for outcomes of ischemic and hemorraghic stroke using feed-forward neural networks

Abstract

Stroke is one of the leading causes of mortality and long-term disability worldwide, with ischemic and hemorrhagic strokes being the two primary subtypes. Accurate prediction of stroke outcomes is crucial for early intervention and improved patient management. In this study, we develop a predictive risk model using a Feed-Forward Neural Network (FFNN) to classify and assess risk factors associated with ischemic and hemorrhagic strokes. The model is trained on a dataset consisting of clinical, demographic, and physiological variables to distinguish between stroke subtypes and predict patient prognosis. Performance is evaluated using accuracy, sensitivity, specificity, and the area under the ROC curve (AUC-ROC). The results demonstrate that the FFNN model achieves high predictive accuracy (95.87%) for training and (80%) for testing in classifying stroke types and estimating risk. This study highlights the potential of deep learning techniques in enhancing stroke risk assessment and decision-making in clinical practice.