Feasibility Study on Fabric-Sheet Unified Sensing Electrode for Non-Contact In-Bed Measurements of ECG, Body Proximity and Respiratory Movement ^†

Abstract

In a recent article, we have reported that a fabric-sheet unified sensing electrode (FUSE) can detect electrocardiogram (ECG), body proximity at the chest and abdomen (BPx_chest, BPx_abd) and respiratory movement at the abdomen (RM_abd) via worn clothes and bedsheet simultaneously from recumbent human subject. We conducted feasibility study involving 21 subjects to assess the influence of individual difference on this FUSE system. Results revealed that individual difference of physical constitution had little impact on both accuracies of ECG R wave and RM_abd under the current FUSE configuration, but had some influence on the voltage changes in BPx_chest and BPx_abd caused by postural change. We deemed to revise some dimensions of the FUSE to avoid individual threshold setting for sitting up detection.

1. Introduction

With the rapid population aging [1], and the progress of information and communication technology, physiological and behavioral signals are considered to be monitored at home in the near future. Vital monitoring at home is considered to play an important role in both preventive daily healthcare and in-hospital care. In a recent research report, a fabric-sheet unified sensing electrode (FUSE) can detect electrocardiogram (ECG), body proximity at the chest and abdomen (BPx_chest, BPx_abd) and respiratory movement (RM) via worn clothes and bedsheet simultaneously from recumbent human subject [2]. In the article, however, verification of principle of the system and fundamental evaluation were only done. In the current research, we conduct feasibility study involving increased subjects to assess the influence of individual difference on this FUSE system.

2. Materials and Methods

2.1. Fabric-Sheet Unified Sensing Electrode (FUSE)

The FUSE can simultaneously measure capacitive electrocardiogram (cECG), BPx_chest, BPx_abd, and respiratory movement at the abdomen (RM_abd) signals. The FUSE is laid under the bed sheet and the back of the lying subject. To prevent sleep disturbance due to night sweat, the FUSE is constructed from breathable materials (conductive fabric of porous copper thin film and insulating fabric; see Figure 1). The detection circuit is basically similar to the conventional system. In addition, a driven seat ground (DSG) circuit [3] was introduced to suppress common mode noise.

2.2. Simultaneous Measurements of cECG, BPx_chest, BPx_abd and RM_abd at Rest and Sit-up Motion

The content of experiment was approved by the Human Life Ethic Committee of Tokyo Denki University. All subjects provided informed consent prior to participation in our experiments. We measured the cECG, BPx_chest, BPx_abd, and RM_abd signals simultaneously totally 31 times from 9 males and 12 females (BMI: 16.0 to 29.0 kg/m²). Subjects were asked to keep supine posture and breathing at 30 breaths / minute (rpm) using the metronome on the bed for 5 minutes. Subjects were then asked to sit up. Reference electrocardiogram signal (ECG_ref) was simultaneously measured by a commercially available wireless telemetry type electrocardiograph (i.e., BN-RSPEC, BIOPAC Systems, Goleta, CA, USA) and disposable electrodes (i.e., F-150S, Nihon Kohden, Tokyo, Japan). Reference RM_abd signal (RM_ref) simultaneously measured by a belt type respiratory transducer of the same telemetry type (i.e., BN-RESP-XDCR, BIOPAC Systems, Goleta, CA, USA). For evaluation, commercially available analysis software (i.e., AcqKnowledge 4.1, BIOPAC Systems, Goleta, CA, USA) was used. Five minutes of rest state was set as an analysis section of the cECG and RM_abd signals. The cECG and ECG_ref signals were preprocessed by smoothed by the 20-point window, filtered through a digital high-pass filter (infinite impulse response filter, f_c = 10 Hz) and differential processing, respectively. The each ECG were detected the R wave by the peak detection function of the software and calculated R-R intervals (RRIs). Detection was performed when the time difference between the RRI of each ECG was ±5 ms or less, and the detection accuracy of each was calculated. RM_abd and RM_ref signals were preprocessed by filtered twice through a digital band-pass filter (infinite impulse response filter, f_c = 0.15–0.5 Hz). Discrete RM_abd cycles in each RM_abd signals were automatically detected by software with threshold processing [4], and detection accuracy of each RM_abd was calculated. Detection of the sit-up motion was calculated from the voltage change of BPx_chest and BPx_abd signals (ΔBPx_chest, ΔBPx_abd) before and after sit-up motion.

3. Results and Discussion

Table 1. Max, Min, MN and SD of the accuracies of cECG R wave and RM_abd signals, and each ΔBPx.

	accuracy of cECG R wave (%)	accuracy of RMabd (%)	ΔBPxchest (V)	ΔBPxadb (V)
Max	100.0	100.0	4.0	6.0
Min	92.5	40.3	0.2	0.2
MN ± SD	97.8 ± 2.2	91.9 ± 12.9	2.1 ± 1.0	3.0 ± 1.6

shows the maximum value (Max), the minimum value (Min), the mean value (MN) and standard deviation (SD) of the accuracies of cECG R wave and RM_abd signals, and each ΔBPx. Figure 2 shows representative simultaneous measured waveforms of BPx_chest, BPx_abd, cECG, ECG_ref, RM_abd and RM_ref signals including sit-up motion. Detection accuracy of cECG R wave exceeded 90% in all subjects. The average value of detection accuracy in RM_abd signal was 91.9%. The ΔBPx due to the sit-up motion varied from 0.2 to 4.0 V in ΔBPx_chest and from 0.2 to 6.0 V in ΔBPx_abd. Measurement of cECG signal were stable in all subjects. On average, the detection accuracy of the RM_abd signals exceeded 90%. Individual difference was confirmed in ΔBPx_chest, and ΔBPx_abd due to sit-up motion. It was inferred that this is due to the difference in bonding position between the electrode and the body due to the difference in physique. It is considered that examination of electrode size is necessary to reduce the influence of difference in body size.

4. Conclusions

Feasibility study was conducted to evaluate the influence of the individual difference on the FUSE. Totally 31 measurements were done for 9 males and 12 females (BMI: 16.0 to 29.0 kg/m²). Individual difference of physical constitution had little impact on both accuracies of cECG R wave and RM_abd under the current FUSE configuration, but had some influence on the BPx_chest and BPx_abd caused by sit-up motion. Therefore, we deemed to revise some dimensions of the current FUSE to avoid individual threshold setting for detecting sit-up motion.