Abstract
Despite many years of research and a few review articles (McAdams et al 1996, Franks et al 2005), the problem of detecting bioelectric signals (ECG, EMG, EEG, etc) from the skin of human subjects has still not been solved in a fully satisfactory way, especially for the case of EMG where signals may be acquired during motion and with large arrays of small and close contacts, making the requirements more challenging. The classical approach involves a small metal surface (10–200 mm2) applied to the skin using conductive gel or glue as an interface (McAdams et al 1996). Capacitive electrodes, dry electrodes, and electrodes with rough surfaces, porous or covered by biolayers or tiny pins (0.2–0.4 mm long) penetrating the 'stratum corneum' are being investigated in many research laboratories (Oehler et al 2008).
The gel used in traditional electrodes has multiple objectives: a) to provide a fluid material that will surround and enclose hairs, therefore providing contact with the skin when the electrode is applied to a hair covered surface, b) to provide an interface that allows minor movements without loss of contact, and c) to keep the skin wet and conductive. Conductive gels are commercially available for ECG and EEG applications. The composition of the gel is sometimes indicated on the container but often it is not. Many commercially available electrodes are pre-gelled with materials whose composition and properties are generally unknown to the user.
The recent development of EMG electrode arrays (high density EMG) including dozens to hundreds of small contact points, each with a surface of about 4–10 mm2 and inter-electrode distances of 5–10 mm, has brought up new problems of distributing conductive gels into many small cavities by dispensers or by spreading with a spatula (Merletti et al 2009). In turn this identified the need to prepare gels that would a) be conductive enough to allow contacts with low and similar impedances, preferably frequency independent in the range 10–500 Hz, b) generate a noise level at gel–metal and gel–skin interfaces no higher than that of the electronics, possibly below 1 μVrms (Huigen et al 2002), c) be dense enough to prevent leakage and short circuits between nearby electrodes but fluid enough to allow filling small cavities without trapping air bubbles, d) be washable in running water in the case of reusable electrode arrays, e) be absorbed by the skin in order to bridge the high resistance of the stratum corneum, and f) show good mechanical stability of the contact in dynamic conditions (Roy et al 2007). As often happens in engineering, some of these requirements are in conflict and compromises must be considered. Despite their many disadvantages, capacitive electrodes may present some useful features to bypass these problems.
Other issues associated with skin treatment (washing, rubbing with solvents or conductive abrasive gels, shaving, stripping, the use of penetration enhancers, etc) are also to be considered. A new area of conductive gel design and skin treatment is developing. Some of these issues are relevant in electrical impedance tomography as well, but in general the levels and frequency ranges are much lower (typically a few pA to hundreds of nA and 10–500 Hz) and the requirements concerning artefacts, stability in time and in conditions of shock and vibration, for long term recording, are more stringent. The issue has been addressed in a small number of articles (Hewson et al 2003, Roy et al 2007, Burbank and Webster 1978), a fact that contrasts with the abundant literature concerning electrodes for stimulation or electrical impedance tomography.
The three articles that follow this editorial provide a preliminary approach to some of these problems. The two works by Freire et al (2010) and Alexe-Ionescu et al (2010) partially address these problems in a frequency range that extends to the electrical impedance tomography bandwidth. The third work (Spinelli and Haberman 2010) deals with an alternative approach of capacitively coupled electrodes.
The work of Freire et al (2010) addresses the issue of gel characterization and investigates the properties of four commercially available gels and five creams using a commercially available stainless steel electrode cell and a novel transparent quartz cell with silver electrodes and a guard ring. An interesting result is the large variability of conductivity (a factor of 40–60 between the lowest and the highest) of the gels and also of the parameters of the constant phase element (CPE) used to describe the metal–gel interface. The commercially available gels are not equivalent.
The work of Alexe-Ionescu et al (2010) focuses on gels based on hydroxyethylcellulose (HEC) and KCl (or NaCl) and investigates the effect of the electrode metal (Ag or Au) and of different concentrations of HEC and salt on the electrical properties of the metal-gel-metal cell.
Although these effects are qualitatively well known and the findings are not surprising, the systematic, quantitative and rigorous approach provided in these articles is filling a gap and should be extended in the future to other gels and to the electrode-gel-skin system with rigorous investigation of other gel materials of different viscosities, different skin treatments, the use of penetration enhancers and of keratin breaking substances. The long term (over hours) osmotic effects of non-isotonic salt concentrations should also be investigated.
The state of the art of capacitive electrodes is reviewed in the work of Spinelli and Haberman (2010). These electrodes require no gel or skin treatment and allow detection of biopotentials through thin clothes. On the other hand, they present high movement artefacts, are vulnerable to power line interferences and require sophisticated shielding and guarding of the front-end stages. They are being studied mostly for ECG detection. Their applicability for EMG electrode arrays (where the many interface impedances must be low and similar to each other and with very low cross-coupling between nearby electrodes) remains to be investigated.
In conclusion, despite the availability of commercial products, and the large amount of empirical knowledge, the issue of optimization of the electrode–skin interface for stable and low-noise bioelectric signal detection (with particular reference to EMG electrode arrays) is far from being solved. The three special section articles that follow provide a preliminary approach that should be expanded by more extensive investigations.
Roberto Merletti, Laboratory for Engineering of the Neuromuscular System, Department of Electronics, Politecnico di Torino, Italy Guest Editor
References
Alexe-Ionescu A-L, Barbero G, Freire F C M and Merletti R 2010 Effect of composition on the dielectric properties of hydrogels for biomedical applications Physiol. Meas. 31 S169–82
Burbank D P and Webster J G 1978 Reducing skin potential motion artefact by skin abrasion Med. Biol. Eng. Comput. 16 31–8
Franks W, Schenker I, Schmutz P and Hierlemann A 2005 Impedance characterization and modeling of electrodes for biomedical applicationsIEEE Trans. BME 52 1295–302
Freire F C M, Becchi M, Ponti S, Miraldi E and Strigazz A 2010 Impedance spectroscopy of conductive commercial hydrogels for electromyography and electroencephalography Physiol. Meas. 31 S157–67
Hewson D J, Hogrel J-Y, Langeron Y and Duchene J 2003 Evolution in impedance at the electrode–skin interface of two types of surface EMG electrodes during long-term recordingsJ. Electromyogr. Kinesiol. 13 273–9
Huigen E, Peper A and Grimbergen C A 2002 Investigation into the origin of the noise of surface electrodes Med. Biol. Eng. Comput. 40 332–8
McAdams E T, Jossinet J, Lackermeier A and Risacher F 1996 Factors affecting electrode–gel–skin interface impedance in electrical impedance tomography Med. Biol. Eng. Comput. 34 397–408
Merletti R, Botter A, Troiano A, Merlo E and Minetto M A 2009 Technology and instrumentation for detection and conditioning of the surface electromyographic signal: state of the art Clin. Biomech. 24 122–34
Oehler M, Schilling M, Ling V, Melhorn K 2008 A multichannel portable ECG system with capacitive sensors Physiol. Meas. 29 783–93
Roy S H, De Luca G, Cheng M S, Johansson A, Gilmore L D and De Luca C J 2007 Electro-mechanical stability of surface EMG sensors Med. Biol. Eng. Comput. 45 447–57
Spinelli E and Haberman M 2010 Insulating electrodes: a review on biopotential front ends for dielectric skin–electrode interfaces Physiol. Meas. 31 S183–98