Analysis And Comparison Of Ventricular Cardiac Arrhythmia Classification Using Sodium Potassium Pump Channel Parameters With ANN And KNN Classifier

Abstract

Aim: The intent of this research is to analyze and compare ventricular cardiac arrhythmia classification using sodium potassium pump (Na+/K+) channel parameters with Artificial Neural Network (ANN) and K-Nearest Neighbor (KNN) classifiers. Materials and Methods: P.J.Noble and A.V.Panfilov model (PJAV) is used for human ventricular study based on the action potential distance. PJAV uses alternative methods of computer simulations which include major ions, pumps and exchangers. Sample was calculated by keeping threshold 0.05, G Power 80%, confidence interval 95% and enrolment ratio as 1. Number of samples considered is 20. According to these data the accuracy is obtained from the classifiers by training novel ANN and KNN classifiers by alternating the Cross fold validation to obtain 20 different samples. These samples are imported to Statistical Package for the Social Science (SPSS) software for graphical representation and overall accuracy. Result: The concluded results shows that ANN has accuracy of 12.25% with standard deviation (4.0911) and Standard error mean (0.9148). Similarly KNN produces an accuracy value of 4.54 % with standard deviation (2.5732) and Standard error mean (0.5754). Conclusion: As of the results, it clearly shows that ANN has better accuracy for classification than KNN.