| Multielectrode Arrays |
|
|
|
![]() Sketch of a typical Axion array. The grid of electrodes, which are 30 µm in diameter and 200 µm apart, provide concurrent access to single cell and network level activity for electroactive tissues and cultures that are introduced onto the array. The utility of an MEA lies in its ability to concurrently monitor single-cell and network-level activity, for extended periods of time, with virtually no destructive interference to the tissue being investigated. In fact, the broad access to network information along with the minimally invasive nature of the device is precisely what makes the MEA an exceptional whole-cell and network-level research tool. However, in order to make MEA systems suitable for high throughput investigations, new technology is required to reduce the cost of the MEA electrodes and to increase the functionality and scale of MEA electronics. The Axion solution to these challenges is described below. Electrodes A major hurdle in MEA research is the expense of the sensor array. The high cost of MEA fabrication arises from the expense incurred in connecting micron-sized electrodes to millimeter-sized sockets and pads for electrical processing (what is known as the packaging problem). The technologies that are employed at the micro-scale (microfabrication) are not economical for millimeter-size features, while the ones used at the millimeter scale cannot produce the required micron-size features. Traditionally, microfabrication derives its economical advantage from the production of hundreds of micro-scale units in parallel. For the fabrication of MEAs, this advantage is lost when the size of the units to be fabricated is drastically increased. If smaller units are built instead, these have to be re-packaged to bridge the size gap, which again increases the costs. To compound the problem, a multi-well platform, such as the Axion Muse-768, requires the packaging of not one, but dozens of micro-scale MEAs in a single unit. Thus, providing more MEAs while reducing the cost-per-unit presents significant challenges in microfabrication. Axion’s patent-pending techniques have solved the scaling and packaging problem simultaneously by performing micro-scale post-processing directly onto a millimeter-scale package. These techniques, illustrated below, allow Axion to reduce cost while dramatically increasing functionality. ![]() A flex-rigid PCB board serves as both the MEA substrate and package (Top Row). Microelectrodes are patterned on top of the PCB substrate (Bottom Row). A central port or hole provides optical access to a transparent laminate membrane that sup-ports the electrodes. Patent Pending. MEA Electronics
High Level channel schematic for Axion’s custom neural interfacing IC. This IC enables simultaneous stimulation and re-cording, and is at the heart of the Axion MEA systems. The current generation of ICs (Achk III-64) services 64 electrodes, providing low-noise amplification, variable filtering, channel blanking, simultaneous sampling, signal multiplexing, current and voltage stimulation, and advanced artifact elimination capabilities, all within a single IC. Further, the AChk III-64 allows up to 4 different current/voltage pairs for stimulation, which can be applied simultaneously to all 64 IC channels. The expandable architecture of the Achk III-64 allows multiple ICs to be used in parallel and forms the core of the Axion Mini and the Axion Muse systems. Artifact elimination Axion’s initial technology for simultaneous stimulation and recording for in-vitro studies was born out of research at Georgia Tech, which began with an inquiry into the source of the stimulation artifact. The intention was to develop a more comprehensive solution to the stimulation artifact problem through a better understanding of the underlying causes. Georgia Tech researchers found that for simultaneous stimulation and recording (i.e., recording on the stimulus channel), the main source of the artifact is the trapped charge that accumulates on the electrode. The four order of magnitude disparity between stimulation voltages and re-cording voltages hinders any open-loop charge-balancing attempt. Even minimal residual charge would saturate the sensitive recording circuitry, which not only causes a prolonged loss of information but also produces extended transients in down-stream filters. Non-linear processes resulting from redox reactions at the electrode-media interface as well as effects from the intervening circuitry further aggravate the artifact and remove any possibility of pre-ordained charge balance. Thus a solution was required that could account for linear and non-linear electrode effects as well as circuit effects. To address the compounding and highly sensitive artifact problems, Axion’s engineers designed an integrated circuit (IC) that places the electrode into an active charge-sensitive feedback loop (US Patent No. 20070178579). This electrode charge control circuitry (EC3) can actively discharge the electrode in a manner that controls both linear charge accumulation as well as transient and non-linear electrode effects. Further, to compensate for higher order effects, circuit-artifacts are prevented at multiple sources, which involves the following strategies or phases: (1) common-mode switching, (2) amplifier blanking, (3) filter response management, and (4) soft switching.
![]() Axion’s IC reduces the artifact to 2 ms or less on the stimulating electrode and 500 µs or less on neighboring electrodes (3200 X gain; BW = 10 hz to 10 khz), and evoked cellular responses have been consistently observed on the stimulating electrode. (Left) Electrode recordings during and after stimulation for various durations of electrode discharge. The gray band indicates the time of an applied stimulus to the media; the stimulating electrode is able to record sine waves applied to a bath of physiologic saline within 2 ms of the stimulus (Blum 2007). (Middle) Cultured cortical recordings with (b) and without (a) artifact elimination (scale bars: 100µV, 10ms, stimulus ±0.5V) (Nam 2008). (Right) Neural recordings recovered by the artifact elimination circuitry on both the stimulating and neighboring electrodes (Nam 2008).
|
| Last Updated ( Monday, 15 December 2008 14:29 ) |






MEAs require specialized electronics to functionally interface to neural tissues and cultures. Typically, neural interfacing technologies for in vitro MEAs use off-the-shelf, discrete, electronics to stimulate and record from MEA electrodes. Unfortunately, the use of discrete components presents some significant drawbacks: (1) it requires vast numbers of chips and components to service a small number of electrodes, (2) it places design constraints on the manner in which the electronics can manipulate the electrode, and (3) it adds tremendous expense to the hardware. Custom designed integrated circuits (ICs) address all of these issues and provide a means to add greater functionality to MEA electronics. Axion has developed circuitry and algorithms that dynamically control the electrode–electrolyte interface to recover neural information that would otherwise be lost to the stimulation artifact. Implementing these technologies in custom ICs enables the economical use of hundreds of electrodes in a manner that would otherwise be impossible. Axion’s ICs include both (1) an electrode control circuitry that interfaces each of the individual electrodes or channels and (2) a scalable system architecture that provides access and control for each electrode. The electrode control circuitry directly interacts with the electrode and provides the means to stimulate, record, and regulate the charge at the electrode–electrolyte interface. The system architecture comprises digital circuitry to coordinate individual electrode functions—such as stimulation—as well as analog circuitry, to condition the signals and provide ‘outside world’ connectivity.
