Seeing Is Believing? A Study of Signal Distortion Produced by Commercial Cortical Microelectrode Recording Elements

Anderson, W.S.

doi: 10.1227/01.neu.0000428420.74457.5f
Science Times

    Phase II invasive epilepsy monitoring is often used to investigate seizure localization and cortical function in patients being considered for surgical resections to help control their seizures. Typical subdural recording elements consist of grids and strips with 1 cm separation between the actual metallic contacts, which themselves are a few millimeters in diameter. Serving both clinical and research contexts, manufacturers of these devices have begun offering versions of the recording systems incorporating microwire arrays which brush against the pial surface and may provide different markers of temporal or spatial seizure onset within the recorded field potential spectral distributions or by gross timing measures of the recorded activity. The microwires have diameters on the order of 40 to 70 µm with spacings between the wires on the order of 1 mm. But can these recording elements be used with existing clinical EEG acquisition systems? Stacey et al (Stacey WC, Kellis S, Patel PR, Greger B, Butson CR. Signal distortion from microelectrodes in clinical EEG acquisition systems. J Neural Eng. 2012;9(5):056007.) now demonstrate in a novel combination of neural simulation and precise impedance measurements that routine clinical systems should likely not be used with these devices because of the profound signal distortion introduced.

    Stacey et al examined over 100 commercially made macro and microelectrodes sold for intracranial recordings. These were examined using the technique of electrochemical impedance spectroscopy to parametrize the complex electrode impedance as a function of applied signal frequency. With this information, and knowing the electrical recording characteristics of typical clinical EEG equipment, the authors developed digital filters or models of the equivalent electrical circuits derived from connecting the recording elements to the clinical recording systems. Using a previously validated neural simulation model of brain activity, the authors then fed the same modeled simulated cortical electrical activity into the digital filters derived from several combinations of micro or macro contacts and recording systems. The authors examined specifically how these realistic neural signals were distorted by the electrode-amplifier combinations.

    In general, macroelectrode systems combined with typical clinical EEG recording instrumentation performed very well with high fidelity in the reproduction of signal waveforms. This is primarily because of the low impedances that macroelectrode systems have. However, microelectrode recording elements had much higher and more variable measured electrode impedances producing significant signal distortion when used with typical clinical recording equipment. This was especially true with low frequency signal components (<60 Hz), which includes the signal range of much of the seizure related activity recorded with EEG (see the Figure). Distortion was less at higher frequencies, which may have implications for the use of microelectrodes in research involving gamma frequency band measurements in cognitive and epilepsy studies. In general, however, the authors recommend the use of high impedance input amplifiers for practitioners interested in the use of microelectrode systems, with input impedances of at least 1 GΩ which is much higher than most clinical systems.

    In summary, Stacey et al explored the impedance and signal transmission properties of several clinical macro- and micro- cortical recording products. The macro elements (typical subdural grid elements) passed electrocorticographic signals with high fidelity, but the micro elements induced unacceptably high levels of distortion in the low frequency range. Microelectrode recording systems in the context of invasive epilepsy monitoring are becoming more widely used in research contexts in an effort to probe higher frequency signals and tighter spatial resolution. They are also marketed more aggressively by the device companies manufacturing them. Neurosurgeons and neurologists interested in implementing these systems in their clinical EEG laboratories would be best served by investing in high input impedance head stage amplifiers or recording systems to best understand the derived recordings.

    Copyright © by the Congress of Neurological Surgeons