Electrical characteristics of conductive yarns and textile electrodes for medical applications
- First Online:
- Cite this article as:
- Rattfält, L., Lindén, M., Hult, P. et al. Med Bio Eng Comput (2007) 45: 1251. doi:10.1007/s11517-007-0266-y
- 758 Views
Clothing with conductive textiles for health care applications has in the last decade been of an upcoming research interest. An advantage with the technique is its suitability in distributed and home health care. The present study investigates the electrical properties of conductive yarns and textile electrodes in contact with human skin, thus representing a real ECG-registration situation. The yarn measurements showed a pure resistive characteristic proportional to the length. The electrodes made of pure stainless steel (electrode A) and 20% stainless steel/80% polyester (electrode B) showed acceptable stability of electrode potentials, the stability of A was better than that of B. The electrode made of silver plated copper (electrode C) was less stable. The electrode impedance was lower for electrodes A and B than that for electrode C. From an electrical properties point of view we recommend to use electrodes of type A to be used in intelligent textile medical applications.
KeywordsTextile electrodesConductive textilesElectrode polarizationElectrode impedance
Biomedical clothing and conductive textiles (also referred to as smart textiles and e-textiles) have in the last decade been of an upcoming research interest. Conductive textiles enable development of smart garments, which are desirable both in health care and military applications, e.g. monitoring of physiological parameters [5, 11–14]. The possibility of knitting or weaving conductive yarns into electrodes, with the advantage of being comfortable and integrated in clothes, make them particularly usable in situations such as long-term monitoring and in home health care [4, 7, 16, 17, 19]. The advantages of such sensors have been thoroughly discussed. However, the performance of such electrodes compared to traditional electrodes is not yet fully investigated.
This study aims to investigate the electrical characteristics of three different conductive yarns (resistance) and three different textile electrodes applied to the skin (impedance and polarization potential) in order to determine their suitability as ECG-electrodes.
By choosing stable electrode materials and suitable gel electrolytes, the difference E and especially the time fluctuation of this difference can be minimized.
When the electrode conveys a current from the body, the ionic current induces a varying potential in the skin electrode with a negative phase angle. The phase shift is due to the charged layer in the transit between the electrode and the electrode gel. The impedance of the electrodes can be described as a resistance in series with a capacitance, where the capacitance varies as C ∼ f−α, where α is a constant .
The electrical properties of three yarns and three corresponding textile electrodes were investigated.
3.1 Yarns and electrodes
3.2 Measurement setup
The first measurement aimed to investigate how the resistance of a single thread varies with thread length, tension and frequency. The resistance of yarns A–C was measured over three lengths (50, 100, and 150 cm) and three tensions (weights 0, 50, and 200 g) and five frequencies (1, 10, 100 Hz, 1 and 10 kHz). Measurements were repeated for two or three samples from each yarn. The investigated yarn was suspended to a rigid support allowing the thread to hang vertically in the air. At the bottom, a weight was applied to obtain a well defined tension in the thread. Sinusoidal signals were fed to the system from a signal generator (model 123, Exact Electronics, Inc, USA). The measurements were made by an Universal Impedance Bridge (Model 250DE, Electro Scientific Industries).
In the second measurement, the polarization potential and electrode impedance for electrodes A–C were measured. The textile electrodes (5 cm2) were moistened with a physiological saline solution (0.9% NaCl). The electrodes were then positioned on the thigh of the test person and held in place with an elastic rib top fabric made out of a non-conducting material. The distance between the electrodes was 5 cm. After each measurement of polarization potential and impedance, the electrodes were moved 5 cm to include effects of local variations in skin properties. Measurements were performed in ten positions for each textile electrode (five on each leg). Standard electrodes (Unilect longterm Ag/AgCl electrodes, Unomedical Ltd, GB) were placed above the knee for reference measurements.
Polarization potentials were measured with a Metrahit 15 s multimeter (ABB Metrawatt, Germany) with input impedance of 10 MΩ by measuring the potential between the two electrodes. The potential difference was registered every minute for 30 min for each position resulting in ten measurement series for each material.
3.3 Data processing
4.1 Thread resistance
4.2 Polarization potential
Figure 5 (right) shows the polarization potential for the textiles and the reference electrodes. The plot shows the median value (when outliers are excluded). The box represents the upper and lower quartiles and the whiskers show the rest of the data content. Outliers (+) were defined as measurements that deviate more than one and a half time the interquartile range from the box. As can be seen in the plot, the results from electrode C show a wide range, while electrode A and B show similar results. The outliers for electrode B, all originate from one measurement series. The reference electrode shows the best results with the lowest median value.
4.3 Electrode impedance
Overall, electrode C has the largest spread of results. Reference electrodes have the lowest impedance but the difference between this electrode and electrodes A and B is limited to a factor of 2.
The phase shift shows capacitive properties which stabilize with increasing frequency. Electrode C has the lowest phase shift, whereas, electrode A and the reference electrodes have the highest.
The usability of a measurement garment is strongly related to the signal quality of the obtained bioelectric signal. As in conventional ECG-recording, the electrodes can be expected to be the weakest link. To obtain adequate electrode properties of smart textiles, the electrochemical factors that determine the electrode potential and liquid junction potential have to be considered in a similar way as for regular electrodes. Furthermore, the theory behind the electrode impedance has to be known. We have shortly presented this theory as a background for the measurements and as a help to interpret the results. The study investigates the electrical characteristics of the yarns and textile electrodes. Electrodes A and B showed performance characteristics close to regular ECG electrodes, but electrode C was considerably worse. Overall, electrode A showed the best electrical characteristics of the textile electrodes.
In ECG-measurements, the electrode impedance is in series with the input resistance of a differential amplifier. A possible adverse effect of a high electrode impedance, is that the suppression of line interferences, decreasing due to a reduction of the effective common mode rejection ratio of the amplifier . Even for monitoring ECG, the line interference can not be totally avoided by filtering since the range of R-wave frequency content coincide with line frequency. An electrode impedance up to 50 kΩ at 50 Hz is probably acceptable . The impedance for electrodes A and B are substantially lower than 50 kΩ at this frequency, while the impedance for electrode C is higher, typically 150 kΩ.
The polarization potential and junction potential must be stabile enough so that the fluctuations do not interfere with the ECG signal. For regular electrodes, this is usually taken care of by appropriate choice of material (e.g. Ag/AgCl) and by suppressing these disturbances using a high pass filter. For example the American Heart Association recommends filtering a diagnostic ECG with a bandpass filter with cut-off frequencies at 0.05 and 100 Hz .
We studied the mean polarization potential drift per minute. If this disturbance comes from low frequencies (with durations longer than 20 s), they will be suppressed by the filter. However, even small variations can pose problems if the frequency coincides with the heart rate. A diagnostic ECG should be able to cover a low rate of 40 bpm. Distributing the polarization potential drift evenly over the heart beats gives an amplitude change of 5, 13 and 40% of 1 mV R-wave amplitude, for electrodes A, B and C, respectively. The corresponding value for the reference electrode is <3%. From this it can be concluded that electrode A is the most stable, B seems to have an acceptable polarization drift even under these worst case circumstances, while electrode C has a possible drift that may disturb the ECG interpretation. The results for electrode C were expected since the polarization potential of silver and copper has poor stability . However, the potential drift might be reduced if the conductive filament is coated with a layer of silver chloride, which can be interesting to consider in further studies.
Our impedance measurement of yarn A–C shows similar results as presented in other studies  and by the manufacturer. Our main contribution in this study is the data on how the yarns react to different tensions and frequencies. Interesting to note is that the yarn B has shown quite high impedance whilst the impedance of the corresponding electrode, electrode B, is comparable to electrode A. A probable explanation is that, when the material is knitted, the number of contact points between the conducting metal staples within the yarn is increased as well as the contact area between the knitted loops, causing a parallel shunting of the current and hence reduces the resistance.
Skin impedance can alter greatly between locations due to varying thickness and composition of the stratum corneum . To include such influences and get consistent measurements, electrodes were placed in different measurement sites. Measurements were performed on the thigh since it allows a relative large area of skin with similar characteristics . The large variation in skin impedance might be an explanation to the presence of outliers presented in Figs. 5 and 6.
Preferably, electrodes should be used without any preparations to increase the usability despite the fact that the perspiration might substantially reduces the impedance. Wijesiriwardana et al.  showed, however that, dry electrodes performed better under long-term conditions than moistened. If the electrode is covered with an impermeable membrane, the body vapor is captured and will function as an electrolyte. Grimnes  showed that it is mainly the perspiration duct that conducts the ions through the skin, resulting in a favorable measurement situation under the presence of perspiration. This is in accordance with a study in the University of Ghent , where a model of the skin- electrode impedance has been investigated under dry and perspiratory circumstances. With our measurement setting and protocol, we observed highly varying impedances for dry electrodes. The final approach, however, included slightly moistened electrodes with a physiological saline solution as a model of a coated electrode.
An approach which has been used to reduce the skin-electrode impedance is to use a membrane of hydrogel as an interlayer between the skin and electrode [2, 13, 16]. The hydrogel has the advantage of being an ionic medium and can at the same time behave as a solid material. The disadvantage is that these membranes eventually will dry out and need to be replaced. On the other hand it is more durable than by merely moistening the electrode. An advantage of the hydrogel is that the electrochemical contact is established instantly and one does not have to wait for the accumulation of sweat.
Impedance measurements have been performed on both yarns and textile electrodes and the polarization potential has been investigated for the electrodes. The yarns showed a purely resistive characteristic and their comparably low resistance enables their use in textile electrodes. The electrode measurements show that both electrode A (multifilaments of 100% stainless steel) and B (staple yarn of 20% stainless steel and 80% polyester) seem to fulfill the required electrode potential stability, where electrode A is the most stable. Electrode C (polyester core with a monofilament of silver plated copper) on the other hand seems not to have the desired stability. In terms of electrode impedance, electrodes A and B have lower values than electrode C. We therefore consider electrode A as the best choice from an electrical properties point of view under the given circumstances.
This work is supported by a fund of the Swedish Governmental Agency for Innovation Systems (VINNOVA).