from Wikipedia, the free encyclopedia

The term neuroprostheses refers to interfaces between the nervous system and the connection to an electronic component for clinical application and medical technology research. Classically, single microelectrodes or electrode arrays (sometimes over 100 electrodes) are used (Santhanam 2006) to correct or restore restricted, pathological or lost functions of the nervous system or to improve normal functions (Rutten 2002, Schwartz 2004 & 2006 ). These technical interfaces can be fundamentally divided into motor and sensory neuroprostheses according to their areas of application . Neuroprostheses are intended to restore failed nerve functions in whole or in part or, as a so-called substitutive method, represent a replacement.

In addition to biological, medical and psychological, the neurosciences now also include philosophical and v. a. also more and more (information) -technological questions. The question of the extent to which the activity of the brain can be artificially simulated by electrical stimulation is even older than the first official findings that precisely those action potentials play a key role in the language of the nervous system ( Galvani 1791, Du Bois 1849). As early as 1755, the physician Charles le Roy attempted to induce visual impressions in blind patients by electrical stimulation of the cortex ( cerebral cortex ) (Le Roy 1755). The electrical stimulation of the cortex was thus one of the first procedures in neuroscience that made it possible to establish a relationship between cortical physiology and perception . The research and medical application of electrical stimulation of neural structures has experienced a revolution in the last few years. V. a. the new biotechnological developments of the interfaces (electrode interfaces) and new mathematical algorithms . The stimulated neuronal populations can be directly activated or also inactivated, which allows a direct investigation of the functional relevance, while recording studies only provide correlations between neuronal activity and perceptual effects (Cohen 2004). The electrical stimulation developed as a widely used method for investigating a variety of neuronal functions of the cellular functioning on perception through to neural plasticity (Maldonado 1996, Ma 2005).

Motor neuroprostheses

The simplest motor prostheses are used to generally promote or inhibit the activity of certain subcortical core areas by applying a carrier frequency. The process called deep brain stimulation has long been used in the treatment of Parkinson's disease or similar basal ganglionic motor diseases. For example, the Ncl. subthalamicus, chronically overactive in Parkinson's disease, inhibited by high-frequency stimulation (Volkmann 2004, McIntyre 2004, Tass 2003). The results of such a brain pacemaker have sometimes been rated very positively by clinical reports, so that in the near future, in addition to other psychomotor diseases such as Tourette's syndrome, classic psychiatric diseases, e.g. B. Depression can be treated with deep brain stimulation (Mayberg 2005). An ambitious goal of future motor prostheses is to enable paralyzed patients to partially restore their motor skills by deriving neuronal signals from the (primarily motor, but not necessarily primarily motor) cortex and converting them into control signals for technical components. Above all, improvements in multi-electrode techniques for deriving signals from entire cell populations have made massive progress possible (Nicolelis 2001, Chapin 2004). These advances have even led to clinical trials on human subjects. These are intended to investigate to what extent neural signals can also be used here to calculate movement trajectories for robotic arms as possible arm prostheses (Nicolelis 2003, Patil 2004, Hochberg 2006, Nicolelis experiment: monkey controls arm). However, it will not be possible to reconstruct the actual fine motor skills of the human hand in the foreseeable future. Current studies in animal experiments have been able to evoke complex motor responses, such as hand-to-mouth movements, defensive reactions or grasping movements, through electrical stimulation of specific regions. So far there has been disagreement as to how the degree of complexity of triggered movements depends on the stimulation location and the stimulation parameters used. Meaning-relevant movements, which are also used in behavioral contexts, have so far been evoked by micro-stimulation from the motor cortex to the premotor cortex to the posterior-parietal or the ventral, intraparietal cortex (Graziano 2002 & 2005, Stepniewska 2005).

Non-invasive, human brain-computer interfaces : A patient with the locked-in syndrome succeeded in moving a mouse cursor on a computer screen and writing e-mails by deriving an EEG by imagining certain movement sequences. Such human, non-invasive brain-computer interfaces (BCI) are also still at an early stage of development. Among other things, Niels Birbaumer from the University of Tübingen is working on such non-invasive human BCIs.

Sensory neuroprostheses


Cochlear implant

The task of sensory prostheses, on the other hand, is to translate physical stimuli into signals that can be used by neurons in order to restore or replace lost sensory functions. Sensory prostheses should enable their users to have meaningfully structured percepts and can start at different levels of the sensory pathways. Here, too, various types of neuroprosthetic approaches are already in the clinical discourse. The only sensory neuroprosthesis that has actually been successfully used therapeutically to date is an inner ear prosthesis ( cochlear implant , CI), which stimulates the eighth cranial nerve (statoacusticus nerve) directly within the cochlea (Rubinstein 1999 & 2001, Middlebrooks 2005). This replaces the mechanical sound transmission via the inner ear and the conversion into an electrical impulse by the hair cells . The quality of the transplant does not come close to the natural hearing sensation. However, the development of this peripheral prosthesis has already progressed so far that a certain group of hearing-impaired to deaf patients can actually understand language, e.g. Sometimes even without lip-reading control (like on the phone), is made possible again (Vandali 1995, Wilson 2003 & 2005, Vandali 2005).


In contrast to the successes in the auditory system, similar developments in neuroprosthetic ophthalmology have so far not been able to offer any clinically used therapies. However, the development of a retina-based prosthesis has made enormous strides in recent years. In this way, much more complex percepts could be generated than simple flashes of light, so-called phosphenes (Humayun 1999, Zrenner 2002, Alteheld 2004, Weiland 2005). German research groups that deal with retinal implants are u. a. in Tübingen under the direction of Eberhart Zrenner . A distinction is made between subretinal and epiretinal implants.

Brain stem implant

Peripheral neuroprostheses, such as the CI and retinal prostheses, can only help a certain group of patients for whom the damage actually occurs before the actual interface (e.g. lack of sound conduction via the auditory bones in the case of inner ear deafness).

To reach another group of patients in recent years, the so-called was auditory brainstem implant ( Auditory Brainstem Implant , ABI) developed. The ABI is a modified CI, whereby it is not the auditory nerve that stimulates directly, but the first switching node of the afferent auditory system, the Ncl. cochlearis in the brain stem (Lenarz 2001, Kuchta 2004, Lenarz 2006, Samii 2007). Current studies by Thomas Lenarz and colleagues deal with another form of central hearing prosthesis in the midbrain ( Auditory Midbrain Implant , AMI). They stimulate specific frequency columns in the superior colliculus with up to now 16 electrodes (Samii 2007, Lim 2007, 2008). However, the technical effort and the clinical risk increases enormously with such central transplants as the ABI. Attempts have also been made for the visual system to evoke visual impressions through direct stimulation of the optic nerve (Veraart) or through direct cortical stimulation on the visual cortex (Dobelle and others). Direct contact with the sensory cortex offers advantages over subcortical strategies. In contrast to, for example, the retina, contact with the biological tissue could be made very stable by using the skull as a fixation for the neuroprosthetic implant. The size and the number of cells in the cortex is much larger and, purely sterically, offers more space for larger electrode arrays, which could be advantageous for restoring the sensory function from an information technology point of view. Furthermore, a broader therapeutic range of disease etiologies could be treated with it, since in principle disorders of the afferent sensory system could be replaced on all levels (Normann 1999, Donoghue 2002 & 2004).

Modern concepts of an interactive, cortical neuroprosthesis

In contrast, it has been shown in a number of human experiments (Brindley 1968, Dobelle 1974 & 1976, Bak 1990, Schmidt 1996) that the generation of visual or auditory percepts is possible in principle by stimulating cortical tissue with electrical current. All of the patients examined consistently reported modality-specific basic impressions , such as phosphene and audene . The assumption that multi-channel implants could simulate the necessary complexity of cortical activity patterns in order to actually create perceptual objects was already expressed around 50 years ago (Krieg 1953). This was followed by work by Brindley, Dobelle and other colleagues on the simultaneous electrical stimulation of many locations in the visual cortex as an approach to a visual prosthesis for the blind. After all, in 1976 Dobelle succeeded in enabling a patient who had been blind for a long time to read Braille letters via such a cortical interface - but only much more slowly than she was able to do tactile. Responsible for this is the low spatial accuracy of superficial stimulation, which does not provide enough independent channels for information. Intracortical micro-stimulation, which can be applied spatially much more specifically, developed from this. But even with this only meaningless and unstructured percepts could be evoked until today (Troyk 2003).

Ultimately, the classic coding approach for cortical sensory neuroprostheses must be viewed as a failure. None of the studies on sensory neuroprostheses since Brindley and Lewin have produced any fundamentally new insights into how complex information can be transferred to the brain with previous stimulation parameters and electrode configurations (Dobelle 2000). Current efforts to explain this failure focus on v. a. to a conceptual further development of the coding approach, which regards the necessary learning performance, in the sense of an application-oriented practice, as a main criterion in order to evoke meaning-relevant cortical stimuli (Scheich 2002). Modern concepts of cortical neuroprostheses could therefore try to combine the biophysical and physiological parameters of electrical stimulation (Butovas & Schwartz 2003, 2006 & 2007, Tehovnik 2006) with the concepts of sensory substitution by Bach-y-Rita (1969 & 2004). An electrical stimulus would then no longer be understood as a restoration of the lost input, but as an abstract, stimulus-coupled input, the meaning of which would first have to be learned. The generation of meaning-relevant percepts is discussed in connection with the philosophy of embodied cognition as a construct of the interplay between body, brain and environment (Chiel 1997, Thompson 2001). This view is supported by many empirical findings. Such a strategy for the development of an interactive, sensory cortical prosthesis in animal experiments is being undertaken by the “Neuroprostheses” working group at the Leibniz Institute for Neurobiology in Magdeburg under the direction of Frank W. Ohl.

Ethical and Philosophical Aspects

The generation or modification of behavior through electrical stimulation of the cerebrum is an ethically and philosophically complex problem. The question of self-causation is already controversial from a physiological point of view ( John Searle 2007), but a study by Talwar et al. (2002), the researchers navigated laboratory rats by implanting three microelectrodes by remote control through even highly complex labyrinths and over supposed dangers - called the Roborat . An electrode innervates the medial forebrain bundle to stimulate the dopaminergic reward system . Two further electrodes simulate the sensory influences of the left and right whiskers in the barrel field of the somatosensory cortex. The reward electrode is paired with the virtual tactile perception left or right and the rat starts moving (Talwar 2002, Nicolelis 2002). The American military is funding this research in the millions. The Pentagon's central research agency (Defense Advanced Research Projects, DARPA ) has launched the Enhanced Human Performance program in order, according to Darpa Director Anthony Theter, “to prevent humans from becoming the weakest link in the US military”.


  • A. Abott: Neuroprosthetics. In Search of the Sixth Sense . In: Nature . 442, July 2006, pp. 125-127.
  • JK Chapin, MAL Nicolelis: Neuroprostheses - robots control with the hand of the mind . In: Spectrum of Science . 1, 2003.
  • A. Krämer: Is the controlled personality coming? In: Spectrum of Science . 9, 2007, pp. 42-49.
  • FW Ohl, H. Scheich: Neuroprosthetics. Hitech in the brain . In: brain and mind . 10, 2006, pp. 64-67.
  • FW Ohl, H. Scheich: Chips in your head . In: Scientific American Mind . April / May 2007, pp. 64-69.
  • H. Scheich, A. Breindl: An Animal Model of Auditory Cortex Prostheses . In: Audiology and Neurotology . 7-3, May / June 2002, pp. 191-194.
  • A. Schwartz: Cortical neural prosthetics . In: Annual Reviews in Neuroscience . 27, 2004, pp. 487-507.
  • I. Wickelgren: Tapping the Mind . In: Science . 299, 2003, pp. 496-499.
  • I. Wickelgren: A Vision for the Blind . In: Science . 312, 2006, pp. 1124-1126.

Web links