Integrated Fuzzy Cognitive Map and Chaotic Particle Swarm Optimization for Risk Assessment of Ischemic Stroke

Stroke diagnosis is an incredibly difficult process since it involves the interaction of both controllable and uncontrollable factors. The diagnosis of stroke is significantly influenced by these factors, which include a variety of factors such as age, blood pressure, gender, obesity, diabetes, smoking, and heart disease, amongst others. It is vital to develop an intelligent system that enables treatment to be administered in a timely and effective manner. This study discusses the application of the soft computing approach, more specifically fuzzy cognitive mapping (FCM), for the goal of estimating the possibility of patients suffering from an ischemic stroke. The chaotic particle swarm optimization technique has been utilized for the purpose of training the FCM training system. The consideration the opinions that were provided by neurologists in order to ascertain the risk rate that was associated with each individual. In order to a cross-validation with tenfold overlap was utilized. The results obtained from this method were compared to those obtained by support vector machine (SVM) and K-nearest neighbour computations, which were performed on 110 real-world observations. The proposed method demonstrated an exceptional level of performance, as seen by its overall accuracy of 94.6 percent and its standard deviation of 3.1 percent.