04-02-2021 | | By Liam Critchey
Neuroelectronic interfaces i.e., brain-computer interfaces, enable the transfer of information from the central nervous system within our bodies to an external device and offer a way of monitoring some of the key neural processes within our bodies.
A lot of research over the years has gone into developing large sensors arrays on flexible materials so that they will conform to the surface they are applied to. This has been a common approach for biocompatible intraneural probes that can map brain activity. This approach has enabled active sensors to become an effective building block for high-bandwidth neural interfaces. They can be developed into these multiplexed sensor arrays that have a large number of neural probes. However, there is a drive within the neural engineering space to make these sensors more effective by increasing their bandwidth, specifically, increasing their sensitivity and spatial resolution (in direct measurement approaches) so that the results produced are more accurate.
Generally, these sensor arrays all have a similar working mechanism. These sensor systems work by modulating the conductivity within a transistor channel. This transistor is electronically coupled with the biological environment via a gate that produces a local signal. This signal can then be amplified to produce a map of brain activity.
Over the years, several different materials have been used to construct the sensing elements and graphene—in the form of solution-gated field-effect transistors (g-SGFETs)—has now emerged as a promising alternative to improve the bandwidth capabilities of neuroelectronic interfaces.
Mapping neurological activity has been no easy task and a range of different materials have been used over the years to build these sensing devices. Like any technology, as new, and more efficient, materials become available, the dynamic shifts, and many options available to scientists and engineers these days.
For a neuroelectric interface to produce sensitive recordings over long periods, they need to have low intrinsic noise, high electrical mobility, high stability, an easy route of integration into flexible substrates, and need to be biocompatible.
For this reason, many semiconducting and semimetal materials have been chosen as the sensing element. For example, organic semiconductor materials and thin silicon nanomembranes have shown a lot of promise already. However, many of the sensors to date have intrinsic limitation (which is why new approaches are constantly being sought). Graphene has emerged as a potential material-based solution alongside some other, more promising nanomaterial solutions tried to date.
Graphene has been gathering a lot of interest in sensor systems for many years now. It is one of the more commercialised areas of graphene-based electronics—with everything from biosensors to Hall-effect sensors (for measuring magnetic fields) being commercially available today. Graphene has been used in sensors, generally, because its high electrical conductivity and charge carrier mobility offer a high sensitivity (small environmental changes can be detected) and its inherent thinness makes it very flexible.
While these properties are useful for accurate and conformal neuroelectronic interfaces, there are several other reasons why graphene has become a material used in these sensor arrays. The high stability (chemical/tensile strength etc.) and general good biocompatibility of graphene are also key characteristics that have played a role in it being used for brain mapping applications. There have been many different sensor arrays created to date—based around graphene FETs (gFETs). Some of these have been highly sensitive to the local biological field. Others have had multiplexed operations, and many have been able to map infra-slow brain activity with a high spatial resolution.
However, many of these high spatial approaches have been via indirect measurements (MRI etc.) due to limited ability to amplify the signal if taken locally as a probe. This is why, even though they have already shown a lot of potential, better graphene neuroelectronic interfaces have been sought that can take direct measurements with a high spatial resolution.
Researchers have now created a new neuroelectronic interface using g-SGFETs. As active sensors, g-SGFETs transduce the electrochemical potential signals in the brain into a drain-to-source current. In these approaches, the signal detection is based on the field-effect mechanism, which helps prevent signal distortion and reduce gain loss. Theoretically, this mechanism could extend to all active FET sensors; however, it has only been realised in gFETs because of graphene’s high stability and chemical inertness.
This new gFET sensing array is composed of flexible 64-channel g-SGFET array and a wireless headstage (to amplify and send the signal to a receiver) and comes from researchers from Germany, Spain and the UK. The sensing array tested on freely moving rodents and was used to wirelessly detect a range of cortical brain signals—ranging from infra-slow to high-gamma frequency bands—for long time periods (up to 24 hours at a time).
The sensing array was developed and tested the way it was to prove that the technology has reached a stage of maturity that it could be used for long time periods, with a high degree of accuracy. The research demonstrated this by not only showing amplification capabilities for direct measurements, but it also illustrated unique features in the infra-slow topographically specific and brain-state invariant patterns. Moreover, the sensors found differences in the infra-slow signals when the rodents experienced slow wave sleep (SWS) and were undergoing rapid eye movement (REM) sleep cycles.
Even though many graphene sensor arrays have been created for mapping brain dynamics, the recent device shows a level of maturity that could enable gFETs to be used an effective tool for mapping epicortical brain dynamics over long time periods (and directly with a high accuracy). This research and the subsequent findings directly impact the field. It showcases how graphene-based technologies are the most suitable candidates for wide frequency band neural sensing interfaces as we advance.