Towards gel-free electrodes: A systematic study of electrode-skin impedance

Abstract

Emerging real-world EEG applications require gel-free electrodes, which have to break through technical bottlenecks and achieve satisfactory electrode-skin impedance. It is crucial to understand the electrical properties of the electrode-skin interface. In this work, the electrode-skin impedance of bioelectrodes (wet, semi-dry, and dry) has been studied systemically, concerning not only magnitude but stability. Various factors have been investigated including types of electrodes, skin locations, pressure, skin abrasion, and electrode contact area. The electrode-skin impedance always decreases in the following order: forearm, scalp and forehead for all electrodes. Compared with the impedances of wet electrodes and semi-dry electrodes, the dry electrode impedances are significantly higher (58.50 ± 64.16 kΩ cm²) and unstable (impedance variation 31.2 ± 31.3 kΩ/10 min). Even worse, the dry electrode impedance variation between six subjects is considerably large (57.5–540.0 kΩ). As a result, no satisfactory EEG signals could be obtained. Moreover, the dry electrode impedances are lowered significantly under pressure or after skin abrasion. Accordingly, alpha rhythms from the dry electrodes appeared with the assistance of pressure or skin abrasion. These findings provide insights for the development of new gel-free electrodes to complement the emerging new EEG applications, such as brain-computer interfaces and wearable EEGs.

Introduction

The brain-computer interfaces (BCIs) have received intense global interest, and the roadmap for the BNCI Horizon 2020 project was launched by the European Commission recently [1]. Under President Obama’s BRAIN Initiative, announced in April 2013, DARPA is currently supporting new research efforts aimed at the development of novel BCI technologies for restoring function in human clinical populations with either neuropsychiatric or memory dysfunction [2]. The Chinse government attaches great importance to brain science researches and the “China Brain Project” is expected to be officially launched by the end of 2016. Moreover, the emerging EEG-based wearable devices have attracted extensive attention in recent years, aiming to real-world scenarios such as physiological monitoring [3], [4], [5], neuro-feedback training [6], [7] and neuro-marketing [8], [9]. However, despite all the recent technological advances in acquisition electronics and signal processing, more practical, convenient and reliable EEG electrodes and their supporting solutions (i.e. headsets, helmets, caps), still remain a crucial technological challenge.

State-of-the-art EEG electrodes are divided into two categories: gel-based electrodes (also referred to as wet electrodes) and gel-free electrodes. The second category can be split further between dry electrodes and semi-dry electrodes. The classical wet electrodes have become “gold standard” in clinics and laboratories for EEG signal recording due to their excellent signal to noise ratio and high reliability. However, the preparation process usually is time-consuming and demands the presence of highly skilled staff [10]. Even worse, conductive gels dirty or damage the hair, and may also cause discomfort for the users [11], [12]. Despite reliable signals, the inconvenience and discomfort issues will severely limit emerging EEG-based applications outside of clinics and laboratories.

To overcome the problem of wet electrodes in real-world scenarios, many efforts have been made to develop dry EEG electrodes in recent years. Dry electrodes consist of an electronic conductor without conductive gel between electrode and scalp, such as inert metallic tips [13], [14], [15], [16], [17], comb-like conductive polymer elastomers [18], [19] and flexible metal-coated bristles [20], [21]. There is no doubt that dry electrodes demonstrate many merits such as cleanliness, quick setup, user friendliness and self-application due to the elimination of conductive gels, which are main concerns for real-world applications.

However, it is noteworthy that, the absence of enough electrolyte always lead to relatively high electrode-skin impedance. Compared to a few kOhms for wet electrodes, the dry electrode impedance is generally in the several hundreds of kOhms or even higher. To the best of our knowledge, the lowest dry electrode impedance is 80 kΩ reported by Cristian et al. [21]. Besides, the dry electrodes, especially rigid electrodes, cannot totally conform to the skin surface, which will cause an unstable electrode-skin interface [22]. As a result, dry electrodes are more sensitive to motion artifacts. All of these cause poor signal quality [23], [24], [25], [26], [27]. An effective approach to improve the signal quality is the integration of preamplification within the electrode [13], [15], [18], [28], [29], [30]. The main advantage of these active electrodes is that, the pre-amplified signal is much less affected by environmental noise [24]. Nevertheless, they are still susceptible to movement artifacts. In addition, active electrodes are usually bulky and expensive [24].

An advance on gel-free electrodes are semi-dry electrodes (or referred to as quasi-dry electrodes) [24], [31], [32], [33], [34]. The semi-dry electrode seems a promising solution for real-word applications because that the concept lies between the classic “wet” and “dry” electrodes, addressing most of their drawbacks, while keeping the advantages of both. The working principle of semi-dry electrodes is to release electrolyte liquid from a reservoir instead of the use of conductive gels. The saline was slowly released into the scalp triggered by electrode adduction under pressure [24], [33], while signal instability was found due to uncontrolled and unexpected liquid release [24]. In order to avoid the above problem, we have proposed a novel ceramic-based semi-dry electrode, which enables recording of excellent EEG signals comparable with wet electrodes [32], [34]. Although semi-dry electrodes demonstrate several merits over dry electrodes, they still need to improve the convenience for users because of the need to supplement saline solution manually [32]. It is clear that there is much progress in gel-free electrodes recently, but there are still a number of problems that need to be addressed.

Electrode-skin impedance is an important parameter reflecting the electrical properties of the electrode-skin interface. The magnitude and stability of the electrode-skin impedance determines the quality of EEG signals. In general, high and unstable electrode-skin impedance can lead to distortions of actual signal [10]. A previous study has shown that no significant attenuation appeared in any of the standard EEG frequency bands when the electrode-scalp impedances reached up to 40 kΩ for clinical applications [35]. In another study, low-frequency noise was found to be increased at high-impedance sites during an oddball task, which caused an increase in the number of trials needed to obtain statistical significance of ERP component [26]. Furthermore, low and stable electrode-skin impedance can minimize the impedance mismatch, which helps to reduce the powerline interference [12]. Last but not least, since motion artifacts are mainly caused by changes on the electrode-skin interface, the electrode-skin impedance is of great importance in estimating effectively the EEG signal quality, especially for real-world scenarios [36]. Generally, the lower the interfacial impedance the more immune the signal will be to movement artifacts. In conclusion, the development of new gel-free electrodes depends on how to maintain low and stable impedance. In other words, the study of electrode-scalp impedance will underpin new electrode advances.

No doubt, in-depth understanding of the electrode-skin interface property by interfacial impedance study is of great significance, which provides workable ideas for new developments of gel-free electrodes and their support systems. The electrical conducting property of the electrode-skin interface from the three typical electrodes (wet, semi-dry, and dry) was quite different, so it’s worth investigating their electrode-skin impedance systematically. However, only very few studies have reported comparisons between these three types of electrodes. In most studies, the impedances were measured at the bare skin or prepared scalp surface with only one or two subjects, which cannot well represent the impedance in practical situations [33], [37], [38], [39], [40]. Although the influence of contact pressure has been studied [41], [42], systematic investigation of various affecting factors on semi-dry or dry electrode-skin impedance have not been reported yet. Other than impedance magnitude, the vital parameter is the stability of impedance that determines the settle time and signal quality. Unfortunately, few studies concerned about the stability of impedance, especially for dry electrodes.

This study systematically investigated the electrode-skin impedance, aiming to provide feasible solutions for gel-free electrodes. The systematic investigation on the electrode-skin impedance focused on comparison of three typical electrodes (wet, semi-dry, and dry). A series of impedance measurements has been carried out including impedance magnitude, short-term stability, and long-term stability. The various factors on impedance have been evaluated including types of electrodes, skin locations, pressure, and skin abrasion. A preliminary study on correlations between dry electrode impedances and EEG signal quality has been made.