Upper threshold phenomenon and its impact on neuroprostheses

Electrical stimulation of neurons is used in the field of neuroprosthetics to artificially generate or block neural activity. Currently the successful example of the cochlea implant is being transferred to the eye to help the blind to regain a rudimentary kind of vision. Stimulation of the spinal cord is used to alleviate back pain or induce movements and control spasticity in spinal cord injured individuals. Modern technology relies on microelectrodes that, instead of penetrating neurons, externally activate them. When a cathodic pulse strong enough to open sodium ion channels on the cell membrane is delivered, the inflow of sodium ions depolarizes the membrane and an action potential is artificially generated. Yet, with even stronger stimuli no action potential is generated or its propagation in one or both directions is prevented. This is called a block or inhibition. Recently, a biophysical principle was proposed that should explain the inhibition of action potentials observed in the soma of neurons in the retina if strong cathodic stimulation is applied with microelectrodes. It was suggested that the strong electric field causes an outflow of sodium ions, which in turn prevents an action potential from forming. It is, however, not clear if this effect of current reversal is of practical relevance, since there is an alternative biophysical explanation. Understanding the implications of these biophysical principles are of importance for both neuroprosthetics as well as neurophysiology. In this project we apply computer simulation to investigate how the geometry of the neurons, their electrical properties and the position of the stimulation electrode influence the inhibition of action potential propagation when strong stimulation is applied.

Since more than 3 decades electrical stimulation of the auditory nerve is a successful method for the deaf or people with severe hearing deficits. This success with cochlea implants is based on the generation of artificially generated nerve signals in the 30000 fibers of the auditory nerve using the comparably quite small numbers of 10-20 microelectrodes, which are able to evoke a pattern of signals good enough for speech understanding. These impressively good results have inspired other applications in neuroprostheses. Examples are inner eye prostheses for visual perceptions for blind people or the activation of muscles for paralyzed persons or after strokes. In such cases where a temporally changing firing pattern within a small group of neurons should be generated it is important to understand the mechanisms involved in order to avoid simultaneously contrary effects. This is of special importance for the retina where about one million nerve fibers transmit the information of a human eye to the visual cortex in man. Currently only 1000 electrode or even much less are in use to enable a rudimentary sight in blind persons. During electrical stimulation with a specific intensity we want to have a better understanding of the mechanisms involved that may activate a group of neurons (retinal ganglion cells) while another group of neurons from the same region is not stimulated as a consequence of a block effect for high stimulation for which the second group is more sensitive. So called ON cells are active for bright light whereas OFF cells of the retina are active when it is dark. A simultaneous activation of both cell types is confusing for the brain as the signals tell it is both bright and dark in the same region. For a better understanding of this situation retinas are stimulated with a microelectrode in animal experiments using a second recording electrode inserted in the cell bodies of retinal cells. The cell bodies are used because of their relatively large size. We could show by simulation of the biophysical events that the observation in the cell bodies cannot guarantee that a silent cell body means no firing in the axon of the cell, a method which seemed to be approved before. Similar relationships were found for electrically stimulated pyramidal cells, the most present neuron type in the brain. In addition, we simulated and analyzed the nerve conduction block mechanisms during electrical spinal cord stimulation for cases of chronic pain. We demonstrated the value of biophysical models and their analysis as an alternative method to animal experiments.