Relationship between Contact Pressure and Motion Artifacts in ECG Measurement with Electrostatic Flocked Electrodes Fabricated on Textile

Abstract

To develop a wearable multi-lead electrocardiogram (ECG) measuring system, we fabricated the electrodes and wires by using electrostatic flocking technology on a textile. By using this technology, it was possible to fabricate many electrodes and wires, simultaneously. Also the flocked electrodes and wires had stretchability and washing resistance properties. To use dry electrodes, it is important to reduce the influence of motion artifacts (MAs). The results of the experiment with the skin phantom revealed that the contact pressure between the skin and the electrode is an important factor in MA reduction. Then, we conducted experiments with a human body to determine the relationship between the contact pressure and the MAs. Under the pressures of 200 Pa and 500 Pa, MAs were observed. Meanwhile, under the pressures of 1000 Pa, 2000Pa and 4000 Pa, the ECG signals under rest and deep breathing conditions were able to be measured without MAs. Considering the comfortability, the contact pressure from 1000 Pa to 2000 is preferable. Finally, we fabricated the wearable ECG measuring system and succeeded in measuring 18-lead ECG signals. The measured ECG waveform is in good agreement with the ECG waveform measured by a commercial system.

Figure 1

Flocked electrodes and wires fabricated by electrostatic flocking technology. (a) A fabrication process of electrodes and wires using electrostatic flocking technology. Ag-plated fiver was flocked on textile with adhesive paste. (b) Photo of electrodes and wires fabricated on textile. (c) Cross section view (A-A′ line) of electrode (d) when skin touched the electrode.

Figure 2

Evaluation of stretchability and wash resistance flocked wires. (a) Mean value and standard error (five samples) of resistance when the flocked wire sample was tensile to 40%. (b) Hysteresis property and recovery property of flocked wire. (c) Washing resistance of the flocked wire and Ag-paste wire. The washing test was conducted according to a standard washing test (JIS L 1930 C4N 140). The graph shows the mean values and standard errors (two samples) of resistance.

Figure 3

Evaluation of flocked electrodes using skin phantom. (a) Image of the experimental setup. ECG signals which were inputted by ECG signal generator and measured by using skin phantom when the contact pressure between skin phantom and flocked electrode was 200 Pa, 500 Pa, 1000 Pa, 2000 Pa, and 4000 Pa. The displacement of the skin phantom was (b) 0 µm, (c) 200 µm, (d) 500 µm, and (e) 1000 µm. (f) Example of the calculation result of correlation coefficients with contact pressure of 500 Pa and with displacement of 200 µm, 500 µm, and 1000 µm. (g) Mean value and standard error (five samples) of the correlation coefficient of ECG signals against master waveform when the contact pressure was 200 Pa, 500 Pa, 1000 Pa 2000 Pa, and 4000 Pa.

Figure 4

Evaluation of contact impedance of the flocked electrodes. (a) Experimental setup to measure impedance while changing contact pressure. (b) Impedance spectra of flock electrode, Ag/AgCl gel electrode, and PEDOT: PSS electrode measured in the 1 Hz to the 10 kHz frequency range in contact with skin (forearm) at 200 Pa. It shows the mean value and standard error (five Samples). (c) Dependence of impedance on contact pressure at the frequency of 100 Hz. It shows the mean value and standard error (five Samples).

Figure 5

Evaluation of flocked electrodes using human body for ECG measurement. (a) A photo in measuring ECG signal when contact pressure was changed and cross-section view and back side of the corset ECG measurement system. ECG signals of the human body (II -Lead) when the contact pressure between the skin and flocked electrode was 200 Pa, 500 Pa, 1000 Pa, 2000Pa and 4000 Pa. The subject was (b) apnea, (c) rest and (d) deep breath. (e) Mean value and standard error of correlation coefficient value of ECG signals against master waveform when the contact pressure was 200 Pa, 500 Pa, 1000 Pa 2000 Pa, and 4000 Pa.

Figure 6

Multi-lead ECG measurement wring and system. (a) Photo of the multi-lead ECG measurement wear. Eighteen electrodes around the chest and four electrodes on lamb were formed by electrostatic flocking technology. (b) Multi-lead ECG measurement system using multi-lead ECG measurement wear and UWB transmitter. (c) Experimental wearable multi-lead ECG measurement system, wireless communication system, and display system.

Figure 7

Experimental result of multi-lead ECG measurement. (a) 18-lead ECG signals measured by the multi-lead ECG measurement system we developed for 5 seconds. (b) One waveform of 18-lead ECG signals measured by the developed system and six ECG signals measured by the ready-made product system.