Selective cell stimulation with retinal implants

An electronic chip implanted in the eye allows blind people to regain vision. New prototypes even enable blind patients to recognize letters and combine them to words. The letters have to be large, the recognition is slow and many mistakes are made. This is because less than 5000 stimulation electrodes replace the approximately 100 million photoreceptor cells in the human eye. Unfortunately, the currently accomplishable resolution is worse than one would expect from the distance between single stimulating contacts on the implants. An analysis of the electrically evoked signals in the retinal cells should improve this situation drastically. During natural (healthy) vision the information process is mainly performed by the retinal cell network that is considered to be kind of a `pre-brain`. The preprocessed signals of photoreceptors are compressed by a factor of 100 before leaving the eye through approximately 1 million fibers in the optic nerve. Finally, the neural signals are processed in the visual cortex where sensory perceptions arise. Most of the first retinal processing is done by cone bipolar retinal cells. This cell type consists of two main groups: (i) ON cells respond stronger to light onset in the center of their receptive field whereas (ii) OFF cells work the opposite way. This counter-play determines information about contrast in the seen picture. However, if both types are activated simultaneously due to an active implant electrode this contrast information gets lost which makes shapes difficult to recognize. This problem takes us to one of the main questions: Is it possible to produce electric potentials that are able to selectively activate ON and OFF cells? A biophysical model analysis will be developed to answer such questions. In a first step, the electric field in the surrounding of a retinal neuron that is generated by a certain electrode geometry gets calculated. In a second step, the response of this nerve cell will be computed. Morphometric cell data and biophysical properties of the cell membrane will be incorporated into a computational model to systematically analyze the neural activity for various cell types, electrode configurations and stimulus amplitudes. The international partner, located at Harvard Medical School, will conduct experiments which investigate the neural response of retinal neurons during external electric stimulation. Besides better knowledge about selective extracellular stimulation, this project will also significantly contribute to the physiology of the healthy eye.

To date, chip-based retinal implants have only permitted a rudimentary restoration of vision. However, modifying the electrical signals emitted by the implants could change that. Two specific retinal cell types respond differently to certain electrical signals - an effect that could improve the perception of light-dark contrasts. Making the blind really see - that will take some time but in the case of certain diseases of the eyes, it is already possible to restore vision, albeit still highly impaired, by means of retinal implants. To achieve this, microchips implanted in the eye convert light signals into electrical pulses, which then stimulate the cells of the retina. One major problem with this approach is that the various types of cells that respond differently to light stimuli in a healthy eye are all stimulated with a common signal. This greatly reduces the perception of contrast. In the research project it was demonstrated to stimulate one cell type more than the other by means of special electrical pulses, thus enhancing the perception of contrast. With the help of a sophisticated computer simulation of two retinal cell types, Frank Rattay and his coworker Paul Werginz showed that by selecting specific electrical pulses, different biophysical processes can actually be activated in the two cell types. For example, monophasic stimulation, where the electrical polarity of the signal from the retinal implant does not change, leads to stronger depolarization in one cell type than in the other. Depolarization means that the negative charge that prevails in cells switches briefly to a positive charge. This is the mechanism by which signals are propagated along nerves.This charge reversal was significantly weaker in the other cell type. Computer simulation showed a fourfold difference in the response of calcium concentrations in the two cell types to a monophasic signal.Calcium is an important signal molecule in many cells and plays a key role in information processing. Concretely, two retinal cell types that are designated as ON and OFF cells were investigated. ON cells react more strongly when the light is brighter at the center of their location, while OFF cells react more strongly when the light is more intense at the edges. The two cell types are arranged in the retina in such a way as to greatly enhance contrast. The problem is that instead of light pulses, conventional retinal implants emit electrical pulses that elicit the same biochemical reactions in both cell types. Consequently, contrast perception is greatly reduced. However, the shape of individual ON and OFF cells affect the way in which the signals are processed. For example, the different length of the two cell types is an important factor. This too, might help to significantly improve the performance of future retinal implants by modulating the electrical signals they emit.