Microelectrode arrays ( MEAs ) or multi-electrode arrays are devices that contain multiple plates or needles through which neuronal signals can be received or emitted. They thus serve as a neural interface that nerve cells can connect to electronic circuits . There are two classes of MEAs: implantable MEAs that are used in vivo and non-implantable MEAs that are used in vitro .
When neurons or muscle cells are stimulated, ionic currents flow through their membranes . This changes charge potentials inside and outside the cell. During recording, the electrodes of the MEA convert the change in potential caused by the ion flow (which corresponds to a voltage change due to a shift in the charge carrier ions ) into a current flow (which corresponds to a shift in the charge carrier electrons ). If the electrodes of the MEAs are used for stimulation , they convert the current flow from the outside into ion flow in the medium. This causes the voltage-activated sodium ion channels of the membrane of an excitable cell to depolarize . This triggers an action potential in a nerve cell and a shortening in a muscle cell.
The size and shape of the recorded signal depend on several factors:
- The type of medium in which the cell or cells are located (e.g. conductivity , capacity and homogeneity ).
- The type of contacts between the cell and the MEA (e.g. area of contact and its density).
- The type of MEA electrode itself (e.g. geometry, impedance ).
- From analog signal processing (e.g. amplification , bandwidth ).
- From digital signal processing (e.g. sampling rate , type of processing).
When capturing a single cell partially covering a flat electrode, the voltage at the contact surface is approximately equal to the voltage of the overlapping region of the cell with the electrode, multiplied by the ratio of the surface area of the overlapped region to the total surface area of the electrode, or:
provided the area around the electrode is well insulated and has a small capacitance. The formula is based on the fact that the electrode, the cell and the surroundings are modeled as a circuit . An alternative tool to predict the behavior of the cell-electrode is a model based on a geometry-based finite element method , which tries to circumvent the limitations by a simplified representation of the system in a compact circuit diagram.
An MEA can also be used to perform electrophysiological experiments on tissue samples or cell cultures . In the case of tissue samples, the connections between the cells within the sample are more or less preserved, whereas in cell cultures they are not present. Neuronal networks form spontaneously in cell cultures of neuronal cells .
It can be seen that the level of voltage that an electrode emits is inversely proportional to the distance between a cell and its depolarization. Therefore, it may be necessary to grow or otherwise place the cells as close to the electrode as possible. In tissue samples, a layer of electrically passive dead cells forms around the incision due to edema . One way of dealing with this is by using MEAs with three-dimensional electrodes that are produced photolithographically . These three-dimensional electrodes penetrate the location of the dead cells of the tissue sample and reduce the distance between the living cells and the electrode. In cell cultures, good adhesion of the cells to the MEA substrate is important for stable signals.
The first implantable arrays were fine wire arrays that were developed in the 1950s. The first experiment to use flat electrodes to record signals from cell cultures was performed in 1972 by CA Thomas, Jr. and colleagues. The structure consisted of an array of 2 × 15 gold electrodes coated with platinum , each 100 µm apart. A cell culture of muscle fibers obtained from embryonic chickens , which was cultivated on the MEA, provided signals with an amplitude of up to 1 mV during the recording. MEAs were built and used independently of the work of Thomas von G. Gross and his colleagues in 1977 to study the electrophysiology of snail ganglia . In 1982 Gross discovered spontaneous electrophysiological activity in cell cultures of spinal cord cells and found that the activity was very temperature-dependent. Below 30 ˚C the amplitude drops rapidly to very small values.
Before the 1990s, laboratories wishing to conduct research with MEAs faced significant entry barriers due to the custom production of the MEAs and the software that they had to develop. However, as cheaper computers and commercial MEA hardware and software became available, many other laboratories were also able to do research with MEAs.
Microelectrode arrays can be divided into subcategories according to their potential applicability: In-vitro and in-vivo arrays.
Types of in vitro arrays
The standard type of in vitro MEA has a grid of 8 × 8 or 6 × 10 electrodes. The electrodes are typically made of indium tin oxide or titanium and have a diameter of 10 to 30 µm. These arrays are used for cell cultures or tissue samples from the brain.
One challenge with in-vitro MEAs was imaging with microscopes that use high-resolution lenses and require a short working distance in the range of micrometers. To avoid this problem, “thin” MEAs have been manufactured that use a covering glass. These arrays have a diameter of about 180 µm, which means that they can also be used with high-resolution lenses.
In another special design, 60 electrodes are divided into a 6 × 5 array that are 500 µm apart. The electrodes within a group are 30 µm apart and have a diameter of 10 µm. Arrays like these are used to study local neural responses and at the same time functional relationships in organic tissue.
The spatial resolution is one of the special advantages of MEAs. It allows signals that are sent over long distances to be recorded with high precision when a high-resolution MEA is used. These arrays usually have a square grid of 256 electrodes, which covers an area of 2.8 × 2.8 mm.
A significantly better resolution is made possible by CMOS-based high-density microelectrode arrays, which have thousands of electrodes with integrated readout and stimulation circuits on compact chips the size of a fingernail. Even the signal propagation along individual axons could be shown. In order to be able to record good quality signals, tissue and electrodes must be in close contact with one another. Punched MEAs apply negative pressure to the openings in the substrate so that the tissue can be positioned on the electrodes to improve the contact and thus the signals.
Types of in vivo arrays
The three main categories of implantable MEAs are: fine wire, silicon-based, and flexible microelectrode arrays.
- The fine-wire MEAs are mainly made of stainless steel or tungsten and can be used to determine the location of individually recorded neurons by triangulation .
- Silicon-based microelectrode arrays contain two special models: Michigan and Utah arrays.
- Michigan arrays allow both a higher density of sensors during implantation and a higher spatial resolution than fine wire arrays. They can also measure the signals along the needles and not just at the end.
- Utah arrays are three-dimensional and consist of 100 conductive silicon needles. With Utah arrays, however, the signals can only be received by the tips of the needles, which limits the amount of information that can be recorded at any one time. They are also manufactured in fixed sizes and parameters, while the Michigan arrays have greater degrees of freedom in manufacturing.
- Flexible microelectrode arrays made from polyimide , parylene or benzocyclobutene have the advantage over fixed microelectrode arrays that they form a closer mechanical connection, since the elastic modulus of silicon is much higher than that of brain tissue, which can lead to inflammation .
Methods of data processing
The 'basic unit' of communication between nerve cells is, at least in electrical terms, the action potential. It is believed that this "all-or-nothing" phenomenon originated on the hilly axon and resulted in a depolarization of the cellular environment that spreads across the axon . The flow of ions through the cellular membrane causes a significant voltage change in the extracellular environment, which the electrodes of the MEA ultimately detect. Therefore, the counting of the voltage peaks and their sorting are often used in investigations for the classification of network activities.
Basically, the great advantage of in vitro arrays compared to more traditional methods such as the patch clamp technique :
- Many electrodes can be used at the same time.
- It is possible to use both experimental electrodes and control electrodes in the same experiment.
- It is possible to select different locations for recording within an array.
- It is possible to record data from different locations at the same time.
- The use can be described as non-invasive as the cell membrane does not have to be pierced, as with most configurations of the patch-clamp technique.
In the case of in vivo arrays, the high spatial resolution is a major advantage compared to the patch clamp technique. With implantable arrays, signals from individual nerve fibers can be recorded, with which information on the position or speed of a motor movement can be recorded. B. a prosthesis can be controlled.
Compared to patch-clamp or dynamic-clamp techniques, in vitro MEAs have a lower spatial resolution. Therefore, they are less suitable for stimulating individual cells. The complexity of the signals that an MEA electrode can send to other cells is low compared to the possibilities of the dynamic clamp technology.
Some biological responses to the implantation of microelectrode arrays are known, especially with permanent implantation. The most important effects are: loss of neural cells, gliosis and failure of electrodes. The response of the tissue to the implantation depends, among other things, on the size of the MEA needles, their spacing, material composition and the duration of the implantation. The response of the tissue is usually divided into a short-term and a long-term response. The short-term response occurs within hours of implantation and begins with an increased number of astrocytes and glial cells around the array. The attacking microglia cause inflammation and phagocytosis of the foreign material begins. Over time, astrocytes and microglia accumulate around the array and form a shell around the array that can extend over several tens of microns. This not only increases the distance between the needles, but also isolates them, which results in a higher impedance being measured. The problems with the permanent implantation of the arrays was the driving force for research on them. A new study examined the neurodegenerative effects of the inflammation that results from permanent implants. Immunohistochemical markers showed the surprising presence of hyperphosphorylated tau proteins , an indicator of Alzheimer's disease , near the recording electrodes. Phagocytosis of the electrode material also raises the question of the biological response. Research indicates that this is low and has virtually disappeared after 12 weeks in vivo . Research into reducing the negative effects of implantation has been carried out on surface coating with polymers such as laminin or with flushable drugs.
In vitro applications
In cultures of neuronal cells, the pharmacological response does not appear to change or diminish compared to in vivo models, so that it can be assumed that MEAs represent a more easily controllable environment for studies in such cultures. So far there have been a number of pharmacological studies using MEAs, e.g. B. Studies with ethanol .
MEAs have already been used as an interface to control non-biological systems through neuronal cell cultures. MEAs can be used as an interface to a neural computer. Cell cultures of rat brain nerve cells were used in a closed control loop to control stimulus-response feedback from an animat in a virtual environment.
A system for a control loop for a stimulus-response model using MEAs was developed by Dr. Potter, Dr. Mandhavan, Dr. DeMarse as well as Mark Hammond, Kevin Warwick , and Ben Whalley at the University of Reading . Approximately 300,000 rat nerve cells were placed on MEAs that were connected to the motors and ultrasonic sensors of a robot that was trained to avoid obstacles.
MEAs were used to study the firing of nerve impulses in tissue samples from the hippocampus .
In vivo applications
There are several implantable interfaces that are currently available to the "end user":
Brain pacemakers have proven themselves in the treatment of movement disorders such as Parkinson's disease . Cochlear implants have helped many people hear better by stimulating the auditory nerve . Because of their remarkable capabilities, MEAs are an important field of research in neuroscience. Research suggests that MEAs can provide deeper insights into processes like memory formation and cognition, and can be therapeutically valuable for conditions like epilepsy , depression, or obsessive-compulsive disorder . In a project called BrainGate (see video at the web links), clinical trials were carried out in which interfaces were used to restore motor control in spinal injuries or as a treatment for ALS . MEAs have the high resolution required to record time-variant signals, making them suitable for both control and feedback of prostheses, as Kevin Warwick , Mark Gasson and Peter Kyberd have shown. Research also suggests that MEAs can help restore vision by stimulating the optic nerve .
- ↑ a b c d e f g h K.-H. Boven, M. Fejtl, A. Möller, W. Nisch, A. Stett: On Micro-Electrode Array Revival. In: M. Baudry, M. Taketani (Eds.): Advances in Network Electrophysiology Using Multi-Electrode Arrays. Springer Press, New York 2006, pp. 24-37.
- ↑ JR Buitenweg, WL Rutten, E. Marani: Geometry-based finite element modeling of the electrical contact between a cultured neuron and a microelectrode. In: IEEE Trans Biomed Eng. Volume 50, 2003, pp. 501-509.
- ↑ a b c S. M. Potter: Distributed processing in cultured neuronal networks. In: Prog Brain Res. Volume 130, 2001, pp. 49-62.
- ↑ a b c d J. Pine: A History of MEA Development. In: M. Baudry, M. Taketani (Eds.): Advances in Network Electrophysiology Using Multi-Electrode Arrays. Springer Press, New York 2006, pp. 3-23.
- ↑ B. Buisson, MO Heuschkel, EM Steidl, C. Wirth: Development of 3-D Multi-Electrode Arrays for Use with Acute Tissue Slices. In: M. Baudry, M. Taketani (Eds.): Advances in Network Electrophysiology Using Multi-Electrode Arrays. Springer Press, New York 2006, pp. 69-111.
- ↑ P. Thiebaud, NF deRooij, M. Koudelka-Hep, L. Stoppini: Microelectrode arrays for electrophysiological monitoring of hippocampal organotypic slice cultures. In: IEEE Trans Biomed Eng. 44, 1997, pp. 1159-1163.
- ↑ a b c K. C. Cheung: Implantable microscale neural interfaces. In: Biomedical Microdevices. 9, 2007, pp. 923-938.
- ↑ CA Thomas, PA Springer, GE Loeb, Y. Berwald-Netter, LM Okun: A miniature microelectrode array to monitor the bioelectric activity of cultured cells. In: Exp Cell Res . 74, 1972, pp. 61-66.
- ↑ D. Eytan, A. Minerbi, NE Ziv, S. Marom: Dopamine-induced dispersion of correlations between action potentials in networks of cortical neurons. In: J Neurophysiol . 92 (3), 2004, pp. 1817-1824.
- ^ R. Segev, MJ Berry: Recording from all of the ganglion cells in the retina. In: Soc Neurosci Abstr. 264, 2003, p. 11.
- ^ A. Hierlemann , U. Frey, S. Hafizovic, F. Heer: Growing Cells atop Microelectronic Chips: Interfacing Electrogenic Cells in Vitro with CMOS-based Microelectrode Arrays. In: Proceedings of the IEEE. Vol. 99, no. 2, 2011, pp. 252-284.
- ↑ DJ Bakkum, U. Frey, M. Radivojevic, TL Russell, J. Müller, M. Fiscella, H. Takahashi, A. Hierlemann: Tracking axonal action potential propagation on a high-density microelectrode array across hundreds of sites. In: Nature Communications. 4, 2013, p. 2181.
- ↑ KJ Angelides, LW Elmer, D. Loftus, E. Elson: Distribution and lateral mobility of voltage-dependent sodium channels in neurons. In: J Cell Biol . 106, 1988, pp. 1911-1925.
- ^ J. Whitson, D. Kubota, K. Shimono, Y. Jia, M. Taketani: Multi-Electrode Arrays: Enhancing Traditional Methods and Enabling Network Physiology. In: M. Baudry, M. Taketani (Eds.): Advances in Network Electrophysiology Using Multi-Electrode Arrays . Springer Press, New York 2006, pp. 38-68.
- ↑ R. Biran, DC Martin, PA Tresco: Neuronal cell loss accompanies the brain tissue response to chronically implanted silicon microelectrode arrays. In: Experimental Neurology. 195, 2005, pp. 115-126.
- ^ GC McConnell, HD Rees, AI Levey, RG Gross, RV Bellamkonda: Chronic electrodes induce a local, neurodegenerative state: Implications for chronic recording reliability. Society for Neuroscience, Washington, DC 2008.
- ^ W. He, GC McConnell, RV Bellamkonda: Nanoscale laminin coating modulates cortical scarring response around implanted silicon microelectrode arrays. In: Journal of Neural Engineering. 3, 2006, pp. 316-326.
- ^ KV Gopal, GW Gross: Emerging Histotypic Properties of Cultured Neuronal Networks. In: M. Baudry, M. Taketani (Eds.): Advances in Network Electrophysiology Using Multi-Electrode Arrays . Springer Press, New York 2006, pp. 193-214.
- ↑ Y. Xia, GW Gross: Histotypic electrophysiological responses of cultured neuronal networks to ethanol. In: Alcohol. 30, 2003, pp. 167-174.
- ↑ TB DeMarse, DA Wagenaar, AW Blau, SM Potter: The Neurally Controlled Animat: Biological Brains Acting with Simulated Bodies. In: Autonomous Robots. 11, 2001, pp. 305-310.
- ↑ SM Potter, R. Madhavan, TB DeMarse: Long-term bidirectional neuron interfaces for robotic control, and in vitro learning studies. Proc. 25th IEEE EMBS Annual Meeting, 2003.
- ^ P. Marks: Rise of the rat-brained robots. In: New Scientist. 2669, 2008.
- ↑ LL Colgin, EA Kramar, CM Gall, G. Lynch: Septal modulation of excitatory transmission in hippocampus. In: J Neurophysiol. 90, 2003, pp. 2358-2366.
- ^ S. Breit, JB Schulz, AL Benabid : Deep Brain Stimulation. In: Cell Tissue Research. 318, 2004, pp. 275-288.
- ↑ K. Warwick, M. Gasson, B. Hutt, I. Goodhew, P. Kyberd, B. Andrews, P. Teddy, A. Shad: The Application of Implant Technology for Cybernetic Systems. In: Archives of Neurology . 60 (10), 2003, pp. 1369-1373.
- ^ AB Schwartz: Cortical Neural Prosthetics. In: Annual Review of Neuroscience. 27, 2004, pp. 487-507.