EMG Measurement with Textile-Based Electrodes in Different Electrode Sizes and Clothing Pressures for Smart Clothing Design Optimization

Abstract

The surface electromyography (SEMG) is one of the most popular bio-signals that can be applied in health monitoring systems, fitness training, and rehabilitation devices. Commercial clothing embedded with textile electrodes has already been released onto the market, but there is insufficient information on the performance of textile SEMG electrodes because the required configuration may differ according to the electrode material. The current study analyzed the influence of electrode size and pattern reduction rate (PRR), and hence the clothing pressure (P_c) based on in vivo SEMG signal acquisition. Bipolar SEMG electrodes were made in different electrode diameters Ø 5-30 mm, and the clothing pressure ranged from 6.1 to 12.6 mmHg. The results supported the larger electrodes, and P_c showed better SEMG signal quality by showing lower baseline noise and a gradual increase in the signal to noise ratio (SNR). In particular, electrodes, Ø ≥ 20 mm, and P_c ≥ 10 mmHg showed comparable performance to Ag-Ag/Cl electrodes in current textile-based electrodes. The current study emphasizes and discusses design factors that are particularly required in the designing and manufacturing process of smart clothing with SEMG electrodes, especially as an aspect of clothing design.

Keywords: EMG; clothing design optimization; clothing pressure; electrode size; electromyography; textile based electrodes.

Figure 1

Electrode preparation in varied electrode diameters (a) and varied pattern reduction rate (PRR) (b). IED = inter-electrode diameter; PRR = pattern reduction rate. Three samples of each condition specification were prepared to reduce the sample error in the comparison.

Figure 2

Actual SEMG electrodes in different diameters (a) and in different pattern reduction rates (b). Bipolar electrodes attached on a leg sleeve were positioned on the rectus femoris (c) and conductive snap fasteners were used to connect to the EMG acquisition system.

Figure 3

Examples of SEMG recordings from one subject during intermittent muscle contraction measured by 5 mm and 30 mm diameter electrodes. (a,b) Show an example of raw EMG signals obtained via electrode diameter Ø 5 mm and Ø 30 mm, respectively. (b) clearly shows lower baseline EMG than (a); (c,d) show filtered (20–500 Hz) in analog and full-wave rectified EMG to analyze the amplitude. Three contractions among five, except the first and last trials, were used to calculate the average activated EMG for comparison.

Figure 4

Average rectified SEMG of baseline and during muscle activation for different electrode diameters. The dashed lines indicate the values obtained from disposable Ag/AgCl electrodes. (a) Baseline electrode noise; (b) SEMG amplitude during muscle contractions exerted by knee extension; (c) Signal to noise ratio (SNR). SEMG was measured on the rectus femoris with stretchable leg sleeves where circle electrodes in different diameters were embedded. Values were calculated by averaging three consecutive measurements from three leg-sleeve samples in each diameter. * p < 0.1 versus 0%, # p < 0.1 versus 10%. Non-parametric statistics verified all significances: the Kruskal–Wallis test was used to identify group differences with the Mann–Whitney U test used as a post hoc test.

Figure 5

Average rectified SEMG of the baseline and during muscle activation for different pattern reduction rates. (a) baseline electrode noise and (b) SEMG amplitude during muscle contractions exerted by knee extension; (c) Signal to noise ratio (SNR). SEMG was measured on the rectus femoris with stretchable leg sleeves where circle electrodes with different diameters were embedded. The values were calculated by averaging three consecutive measurements from three leg-sleeve samples in each diameter. * p < 0.1 versus 0%, # p < 0.1 versus 10%. Non-parametric statistics verified all significances. The Kruskal–Wallis test was used to identify group differences with the Mann-Whitney U test used as a post hoc test.

Figure 6

Average rectified smoothed SEMG during a full-depth squat. The electrodes were embedded on the leg sleeves with a varied pattern reduction rate (PRR) of (a) 0%, (b) 10%, (c) 20%, and (d) 30% so that the muscle activities of the rectus femoris were recorded. The values were averaged from three leg-sleeve samples in each PRR.