An Efficient Classifier-Based Approach for Early Arrhythmia Detection with Feature Reduction Using Ranker Search Algorithm
The increasing rate of sudden cardiac arrest (SCA) is a measure cause of death across the globe. The arrest may too take place with no forewarning; therefore, the primary estimation of “ventricular tachycardia” (VT) and “ventricular fibrillation” (VF) arrhythmia conditions is very much crucial for the precise detection of ventricular arrhythmia conditions. The present work aims at the detection and classification of VT and VF arrhythmias by extracting the temporal, statistical, and spectral features of the ECG signal. Two different databases from PhysioNet repository, i.e., CUDB and VFDB, were considered for evaluation of the proposed algorithms. For a 5 s window length, a total of 19 features were extracted. The ranking of the attributes has been by using correlation attribute evaluation method together with a ranker search algorithm. The normal sinus rhythm (NSR), VF, and VT arrhythmia rhythms are classified by J48 decision tree algorithm, SVM, MLP, and CVR classifiers. A total of 57 records of ECG signals have been evaluated to attain an accuracy of 97.73%, sensitivity of 92.78%, specificity of 98.11%, and precision of 96% using J48 decision tree algorithm, which gives better result compared to other three classifiers. This method may be served as an imperative decision support tool in precise recognition of ventricular cardiac arrhythmia.
KeywordsECG signals Sudden cardiac arrest (SCA) Ventricular tachycardia (VT) Ventricular fibrillation (VF) Features Decision tree SVM
- 5.Alonso-Atienza F, Morgado E, Fernandez-Martınez L, Alberola AG, Rojo-Alvarez JL (2012) Combination of ECG parameters with support vector machines for the detection of life-threatening arrhythmias. Comput Cardiol 39:385–388Google Scholar
- 7.Zhang XS, Zhu YS, Thakor NV, Wang ZZ (1999) Detecting ventricular tachycardia and fibrillation by complexity measure IEEE Trans Biomed Eng 46:548–555Google Scholar
- 8.Karegowda AG, Manjunath AS, Jayaram MA (2010) Comparative study of attribute selection using gain ratio and correlation based feature selection. Int J Inf Tech Know Man 2:271–277Google Scholar