Passive Capacitive ECG Sensing: Assessment of Signal Quality During Different Types of Body Movement

Devices for long-term measurement of the electrocardiogram (ECG) provide means to detect sporadic cardiac arrhythmiae such as atrial fibrillation. While ECG devices with galvanic coupling between body and electrodes have a limited wearing time due to drying out and can cause skin irritation, capacitive coupling is not limited in these respects and can thus increase wearing comfort and measurement time of ECG monitoring systems. For integration into daily life, however, robustness against physical activity is a crucial requirement for such devices. We here evaluate the effect of body motion on ECG signal quality for a recently proposed device.The ECG was measured in seven subjects during phases of movement of different parts of the body and compared to a commercial reference ECG device with galvanic coupling. Data quality was assessed in terms of sensitivity and precision of R-peak detection with the established Pan-Tompkins algorithm.The signals were corrupted by both high-frequency disturbances as well as strong baseline variations, which lead to 33% of R-peaks being missed in average. These disturbances were most pronounced during full-body movement as well as tension of the chest muscles, whereas signal quality was hardly affected during leg movement. During moderate activity, precision ranged between 92% and 70%. Of the false negative classifications 43% in average were due to baseline variations causing sensor overload. Furthermore, a strong inter-subject variation was observed.It is argued that the observed motion artifacts, particularly strong baseline variations, are due to changes in contact impedance of the measurement electrodes. In order to provide usability of the capacitive ECG device in daily life, these asymmetries in contact impedance have to be counteracted, particularly by flexible electrode design.