Potential Risk Factor Analysis and Risk Prediction System for Stroke Using Fuzzy Logic

Stroke is a life-threatening, deadly cause, which occurs due to the interruption of blood flow to any part of brain. As stroke is a globally alarming deadly cause, using computational expertise to aid this problem, is high on demand. In this paper, our proposed system focuses on the potential risk factor for system design. Using computational technique, we prune unnecessary risk factors which are less likely to cause stroke on patient dataset collected from a medical college in Bangladesh. Fuzzy C-means classifier and Fuzzy Inference System are used to classify input data. Later on, to generate fuzzy rule we use Adaptive Neuro-fuzzy Inference System so that it can give better prediction. The developed system provides higher accuracy which satisfies the physicians’ demand. Therefore, the developed system will aid not only general people but also medical experts.