ANNUAL JOURNAL OF ELECTRONICS, 2014, ISSN 1314-0078 77 Methods for Detecting Pacemaker Pulses in ECG Signal: A Review Valentin Viktorovich Tsibulko, Ivo Tsvetanov Iliev and Irena Ilieva Jekova Abstract - This paper reviews some different methods for detecting the pulses created by an implanted cardiac pacemaker in the ECG signal. It includes a hardware method implemented in a Texas Instruments circuit, software method proposed by Herleikson and patented by Hewlett-Packard, an upgrade of Herleikson’s algorithm reported by Polpetta and Banelli and an software pace pulse detection embedded in the analog front end module of Analog Devices – ADAS1000. Keywords – pacemaker, pace pulse detection, ECG I. INTRODUCTION Correct detection of pacemaker pulses in the electrocardiogram (ECG) is crucial for proper evaluation of the effect of a pacemaker on the cardiac rhythm. The ECG signal recorded from a patient with implanted cardiac pacemaker consists of three parts: the natural ECG signal, pacemaker pulses and noise. The informative content of the ECG signal lies in the frequency band (0 – 150) Hz and its dynamic range is usually up to 2mV. On the other hand, the pacing pulses have a typical duration from 0.1 ms to 2 ms [1-3], and amplitude higher than 0.5 mV. They have very fast rising and falling edges - the rising edge duration could be 100ns measured at the pacemaker leads, appearing widened to 10µs on the surface of the human body [1]. The detection of pacing artifacts is important, since they indicate the presence of a pacemaker and help to evaluate the reaction of the heart. There are different medical standards with variable requirements regarding the height and width of the pace pulse that has to be captured and indicated on the screen of the device. According to ANSI/AAMI EC11 [3] the features of the pacemaker pulses that should be obligatory detected are as follows: • duration - 0,1ms to 2ms • amplitude - 2mV to 250mV • frequency - up to 100 impulses per minute • rising edge duration - less than 100ms The IEC60601-2-27 standard [4] states different requirements towards the duration (0.5 ms to 2.0 ms) and the amplitude (2mV to 700 mV) of the pulses. Modern pacemakers could generate smaller pacing pulse amplitudes that could fall below the requirements set in the standards and lead to complications in the algorithms for pacing pulses detection [5]. Another challenge faced by the methods for pacing pulses recognition is the noise part of the signal that consists of: • baseline wander due to the patient respiration and movement - 0.05 Hz to 1 Hz; • power-line interference - 50/60 Hz, and the respective harmonics; • electromyographic (EMG) noise, caused by the electrical activity of the muscles. The baseline wander and the power line interference are with relatively low frequency and do not disturb the detection of the pacing pulses in the ECG. However, the frequency band of the EMG is reported to be up to 5000 Hz [6] and depending on the sampling rate it can overlap with the pace pulses frequencies, thus causing some serious difficulties for the correct pace pulses recognition. A pacing pulse that has passed thought a low-pass filter, can be significantly widened, and when it has passed to a high-pass filter, a tail at the end of the pulse can be created. Considering the above listed problems, the algorithm for pacing pulse detection should be applied on high-resolution ECG [7, 8] that preserves the frequency content of the pacing pulses. The aim of this paper is to present some existing hardware and software methods for pace pulses detection. II. METHODS FOR DETECTION OF PACE PULSES The pacing pulses detection could be hardware, software or a combination of both. Fig. 1. Block-diagrams of hardware and software pace pulse detection methods V. Tsibulko and I. Iliev are with the Department of Electronics and Electronics Technologies, Faculty of Electronic Engineering and Technologies, Technical University - Sofia, 8 Kliment Ohridski blvd., 1000 Sofia, Bulgaria, e-mail: valentin.tsibulko@gmail.com, izi@tu-sofia.bg I. Jekova, is with the Institute of Biophysics and Biomedical Engineering, Bulgarian Academy of Sciences, Acad. G. Bonchev str., bl. 105, 1113 Sofia, Bulgaria, e-mail: irena@biomed.bas.bg