Respiratory Monitoring with Textile Inductive Electrodes in Driving Applications: Effect of Electrode's Positioning and Form Factor on Signal Quality

Abstract

This paper provides insights into where and how to integrate textile inductive electrodes into a car to record optimal-quality respiratory signals. Electrodes of various shapes and sizes were integrated into the seat belt and the seat back of a driving simulator car seat. The electrodes covered various parts of the body: upper back, middle back, lower back, chest, and waist. Three subjects completed driving circuits with their breathing signals being recorded. In general, signal quality while driving versus sitting still was similar, compared to a previous study of ours with no body movements. In terms of positioning, electrodes on seat belt provided better signal quality compared to seat back. Signal quality was directly proportional to electrode's height on the back, with upper back outperforming both middle and lower back. Electrodes on the waist provided either similar or superior signal quality compared to electrodes on the chest. In terms of form factor, rectangular shape outperformed circular shape on seat back. Signal quality is proportional to the size of circular electrodes on seat back, and inversely proportional to size of rectangular electrode on seat belt.

Keywords: automobile; automotive application; breathing sensor; driver health status; driving application; inductive sensing; respiratory monitoring; respiratory signal; signal quality index; textile inductive electrode.

Figure 1

The driving simulator. (a) Complete setup. (b) Zoom-in on the car seat. The textile inductive electrode is hidden behind the black substrate fabric between the two white crosses. (c) The circular textile inductive electrode A.73.12.

Figure 2

Diagram showing textile inductive electrodes’ positions. (a) There are three positions on the seat back: upper back (1), middle back (2), and lower back (3) position. (b) There are four positions on the seat belt: chest (1, 2), and waist (3, 4).

Figure 3

Print screens of the driving simulator’s user interface. (a) Scenery. The road is bordered with greenery, including trees. While driving, the subject is following a car. (b) Driving circuit. The green line represents the road. The black dots are trees. The circuit starts at the red dot, in the direction of the blue arrow.

Figure 4

Examples of respiratory signal with various signal quality. Amplitudes are all in standard score unit and displayed within the same axis limits to allow direct comparison. (a) Good signal quality. Recorded on the chest of subject #1 with a 50 cm² rectangular electrode on seat belt. (b) Average signal quality. Recorded on the chest of subject #3 with a 50 cm² rectangular electrode on seat belt. (c) Poor signal quality. Recorded on the middle back of subject #3 with a 28 cm² circular electrode on seat back.

Figure 5

General assessment of signal quality: reference vs. textile. (a) Comparison of SBR values. (b) Comparison of SHR values. (c) Comparison of MMR values. SQIs are sorted in decreasing order and plotted for reference and textile electrodes. There is one dot for each respiratory signal. The value of one dot is the average of 34 SQIs calculated on 34 one-minute sections of signal.

Figure 6

Signal quality vs. electrode’s position on seat back. Normalized SQMs are sorted in decreasing order and plotted. There is one dot for each one-minute section of textile respiratory signal. The value of one dot is the SQM calculated for one section of respiratory signal.

Figure 7

Signal quality vs. electrode’s size on seat back. Normalized SQMs are sorted in decreasing order and plotted. There is one dot for each one-minute section of textile respiratory signal. The value of one dot is the SQM calculated for one section of respiratory signal.

Figure 8

Signal quality vs. electrode’s shape on seat back. Normalized SQMs are sorted in decreasing order and plotted. There is one dot for each one-minute section of textile respiratory signal. The value of one dot is the SQM calculated for one section of respiratory signal.

Figure 9

Signal quality vs. electrode’s position on seat belt. (a) Chest vs. waist with 50 cm² electrode. (b) Chest vs. waist with 75 cm² electrode. For each panel, normalized SQMs are sorted in decreasing order and plotted. There is one dot for each one-minute section of textile respiratory signal. The value of one dot is the SQM calculated for one section of respiratory signal.

Figure 10

Signal quality vs. electrode’s size on seat belt. (a) 50 cm² vs. 75 cm² on chest. (b) 50 cm² vs. 75 cm² on waist. Normalized SQMs are sorted in decreasing order and plotted. There is one dot for each one-minute section of textile respiratory signal. The value of one dot is the SQM calculated for one section of respiratory signal.

Figure 11

Signal quality: seat back vs. seat belt. Normalized SQMs are sorted in decreasing order and plotted. There is one dot for each one-minute section of textile respiratory signal. The value of one dot is the SQM calculated for one section of respiratory signal.