Feature Selection Algorithm for ECG Signals and Its Application on Heartbeat Case Determining

This study proposes a simple and reliable feature selection algorithm for ECG signals which can be applied on heartbeat case determining. A well-established technique, Principal Component Analysis (PCA), is employed for qualitative feature selection in this study. Determination of heartbeat case is carried out by fuzzy logic and Fisher’s linear discriminant analysis (Fisher’s LDA). The proposed method has the advantages of simple mathematic computations, high speed, low memory space, good detection results, and high reliability. This study consists of three major stages: (i) QRS extraction stage for detecting QRS waveform using the Difference Operation Method (DOM); (ii) qualitative features stage for qualitative feature selection from ECG signals using the PCA method; and (iii) classification stage for determining patient’s heartbeat cases using fuzzy logic and Fisher’s LDA. Records of MIT-BIH database are used for performance evaluation. Experimental results show that the total classification accuracy (TCA) is 94.03% and 93.87% for methods by using fuzzy logic and Fisher’s LDA, respectively.