Determination of the noise source in the electrocardiogram during cardiopulmonary resuscitation

During cardiopulmonary resuscitation (CPR), the electrocardiogram (ECG) is often obscured by noise. This noise is in the form of baseline variations in the ECG, which often necessitate stopping chest compressions to adequately assess the ECG. Because survival from cardiac arrest has been shown to be related to blood flow generated during CPR, and because interruption of chest compressions will reduce blood flow, survival may be compromised by these interruptions. Three possible sources for the noise were considered: the heart, which is deformed during CPR, which may introduce a mechanical-electrical interaction and alter the normal electrical pattern of the heart; the thoracic cavity, which may have large impedance variations because of CPR and thereby modulate the ECG; and the skin-electrode interface, which may be mechanically disturbed during CPR and thus produce polarization potentials that cause additional noise. CPR studies were performed on five dogs by using four different test conditions and six different types of electrodes. The test conditions were: electrode motion, which allowed mechanical disturbances of the skin-electrode interface without altering the thoracic impedance or deforming the heart; vest CPR; manual CPR; and respiration. The myocardial ECG, the bipolar and unipolar surface ECGs, and the thoracic impedance were monitored. Different types of surface ECG electrodes were used to determine whether the noise was dependent on electrode type or size. There were no baseline variations in the myocardial ECG during any of the test conditions. The thoracic impedance did vary during CPR, but the variations were temporally uncorrelated to the baseline variations in the ECG, and the variations were of similar magnitude as the variations caused by respiration, which produced no baseline changes in the ECG. Finally, the magnitude of the baseline variations in the ECG was substantially different for electrodes of different sizes and shapes, and electrode motion produced baseline variations that were identical to those produced during CPR. Therefore, it was concluded that the source of the noise in the ECG during CPR is the skin-electrode interface and, specifically, that the noise is related to the electrical properties of the electrode.