Noise in biopotential recording using surface electrodes

Electric events inside the human body cause potential differences on the skin. Surface electrodes can be used to measure these voltages. A surface electrode is in fact an electrochemical transducer. It converts the potential differences due to ion flows inside the human body into a measurable voltage. The measurement of bioelectric events is exposed to various sources of noise. The reactions that take place at the electrode make the electrode itself a source of noise. This paper is the result of a study of available literature on noise in biopotential recording. It focuses on the noise that originates at the electrode-skin interface. The silver-silverchloride electrode is commonly used in low-noise recording. It has a stable halfcell potential and is nonpolarizable. In a nonpolarizable electrode unhindered charge transfer is possible, making it suitable for stimulation, in a perfectly polarized electrode no charge exchange is possible. When the electrode is applied to the skin the noise level measured is higher than the level that is to be expected from thermal noise only, generated by the resistive impedance of the electrodes and intermediate tissues. The origin of the excess noise is still unclear. Localized out-of-equilibrium processes at the upper skin layers are suggested, as well as a residual EMG signal from nearby muscles. Abrading the skin lowers the excess noise. The noise level is also dependent on the type of electrode used. This suggests that processes at the electrode-electrolyte interface add to the total noise level. Examples of other noise sources are amplifier noise, capacitive and inductive interference and motion artifacts. Low-noise amplifiers are available, and measures can be taken to reduce the noise level of the latter sources to negligible value. This causes the voltages generated at the electrode-skin interface to become the dominating noise source. A capacitive electrode is a conductor coated with an insulating layer. When applied to the skin, the electrode is capacitively coupled to the skin, and its potential is therefore representative for the potential of the skin. The insulating layer has to be very thin, and is therefore vulnerable. Only few examples are mentioned in literature. Little data is available of the actual performance. In a truly capacitive electrode there is no conduction between metal and electrolyte. No halfcell potential will develop. The capacitive component of the electrode is noise-free. Therefore the capacitive electrode can be used for the study of electrode noise processes, as the interaction …