Increasing the Speed and Precision of Prediction of the Results of Angiography by Using Combination of Adaptive Neuro-Fuzzy Inference System and Particle Swarm Optimization Algorithm based on Data from Kowsar Hospital of Shiraz

With regards to the importance of early prognosis of coronary artery diseases, in recent years the use of the latest artificial intelligence and data mining findings is considered to assist physicians. The purpose of this study was to increase the precision and prediction speed for the results of angiography by using a combination of fuzzy inference systems and particle swarm optimization algorithm. Materials & Methods: A new system consisting of a combination of fuzzy inferences and particle swarm optimization algorithm was proposed and simulated by MATLAB software R2015a (8.5.0.197613). The samples consisted of 152 patients who were randomly selected from those undergone coronary artery angiographies in Kowsar Hospital of Shiraz, Iran, in August 2013. The data were then analyzed by Excel 2010 and the essential parameters of the proposed system were extracted. Findings: The data were then randomly divided into 20 groups for training and testing. These groups were selected randomly in a manner that 85% of the data were used for training and 15% for testing, and each group was simulated individually. The results of the simulation after 20 rounds of simulation with different training and testing data in system performance indicators displayed that the average of sensitivity, specificity, precision, and accuracy was 0.8422, 0.9192, 0.8554, and 0.8888, respectively, and it was equal to 1 in the most optimal situations. Discussion & Conclusions: High performance indicators prove that the proposed system has a satisfactory performance in predicting the results of angiography and classifying them into two classes of normal and patient. In fact, in this study, prediction speed and precision were improved by using the proposed system, which was based on neuro-fuzzy inference system in combination with particle swarm optimization meta-heuristic algorithm.