Application of the Kalman filter in cardiac arrhythmia detection

Abstract

The variability of the R-R intervals in the ECG signal contains valuable information about the various types of arrhythmia that might be present. It has been recently suggested, that the identification of cardiac arrhythmia might be possible by applying spectral analysis methods. This paper intends to investigate the efficiency of a spectral analysis method, namely the application of the Kalman filter identifier in the calculation of time varying spectra of the R-R interval time series. The efficiency of the method is tested using the MIT-BIH database and another database built up at the Medical Clinic No. 3 of Tg. Mures; in particular, cases of bigeminy, trigeminy, second degree block and ventricular flutter have been tested. Tests have revealed, that this technique in most cases can detect the onset of arrhythmia and can also identify the arrhythmia that is present.