Non-contact ECG monitoring

Abstract

Listen

Translate article

Comment on `Electric potential probes - new directions in the remotesensing of the human body'Dear EditorThe authors of `Electric potential probes - new directions in the remotesensing of the human body', which appeared in your February issue(2002 Meas. Sci. Technol. 13 163-9), are tobe congratulated on their demonstration of non-contact ECG monitoring.Although they probably overstate the problem of signal distortion whenusing conventional electrodes, a monitoring modality that removes thenecessity for electrode jelly and skin preparation is certainlyadvantageous.However, with reference to their figure 5, the remote off-body ECGtrace, I am interested in the authors' assertion that the signaldisplayed is due to cardiac activity and not movement. Assuming theheart rate of the subject was in the normal range of 60 to 80bpm, theoscillatory signal clearly visible between beats can be estimated atbetween 6 and 8 Hz. To my knowledge a signal component of this frequencycannot be obtained from any combination of surface ECG electrodes.Because of the front-to-back measurement used we might expect thenon-contact signal to resemble a conventional ECG taken between theequivalent areas of the body, which it doesn't.Given that the stated filtering for this signal is 1 to 30 Hz, it isunlikely that this oscillation is a function of the filter (in contrastto their figure 6(b) where the 5 to 15 Hz bandpass filter is beginningto selectively pick out the centre band sinusoidal component).This phenomenon is of interest to me because in recent, unpublished workinvolving measurement of the cardioballistogram (the physical movementof the torso at each heartbeat in reaction to the ejection of blood fromthe left ventricle into the aorta) I specifically recorded dampedlow-frequency oscillations, very similar to those in figure 6, which provedto be a sympathetic mechanical oscillation of the viscera in response tothe cardiac ejection stimulus.My question to the authors, therefore, would be what experiments didthey perform to determine that their signal was, in fact, electricalrather than cardioballistic in origin?John Brydon Consulting Engineer, PO Box 155, Artarmon, New South Wales 1570, AustraliaReply from the authorsDear EditorWe would like to thank Dr Brydon for his interest in our paperand we appreciate his congratulations on our demonstration ofnon-contact ECG monitoring.With regard to his opinion concerning signal distortion, our statements in the paper were based on the observation that our sensors consistently detect larger amplitude surface potentials than conventional gel electrodes. This indicates to us that electrical loading of the body occurs with conventional electrode systems.Dealing with Dr Brydon's second point, we would make the following observations. Because of the capacitive nature of the coupling between our sensor and the body, one would expect any movement artefacts to be most pronounced at small distances from the surface of the body and to decrease with increasing distance. This is not the case. As regards the remote, non-contact signals measured from the front to the back, we do not agree that these should resemble conventional ECG measurements. First, of course, such remote, off-body detection is new - it has not been carried out before. There is therefore no other data available with which to compare our results. Second, the heart is a multi-polar source with distributions of fields and potentials which must vary strongly as a function of distance from the body. One would therefore not expect the ECG at (say) 1 metre away from the body to be the same in detail as that sensed at the body surface. It is clear that at such large distances the signal detected by our sensors is a spatially averaged composite of the electrical activity across the chest.C J Harland, T D Clark and R J Prance School of Engineering and Information Technology, University of Sussex, Brighton BN1 9QT, UK