Home Research BCCN I Sections Section D
Personal tools

Section D

Motor Adaptation and Neuroprosthetics

A most suggestive area for applications of computational neuroscience is found in neuroprosthetics, which directly builds upon knowledge on biologically inspired control algorithms and analysis methods of analysis for neural data. The center has initiated a close cooperation with the company Otto Bock HealthCare, which will be highly advantageous for the application and the exploitation of a substantial part of the scientific work done at the center. Otto Bock HealthCare presently is the only global player in the rehabilitation industry and the world's largest manufacturer of prosthetic and orthetic components and is active in many related branches such as development and fabrication of rehabilitation and mobility products, as well as special materials for such products. The company clearly is one of the leaders in technology in all of its areas and - being the market leader - is able to direct efforts in research and development by long term strategic goals. The projects in section D focus on neural mechanisms underlying the acquisition and execution of motor skills. These mechanisms are studied at the level of basic research or as more application-oriented collaborative projects. There exists unambiguous experimental evidence that the activity of individual neurons in motor cortex is quantitatively related to movements, which thus can be predicted by analyzing data from motor cortical neurons. The more complex adaptation processes that are implied by the task-dependent recruitment of motor neurons, however, behave counter intuitively under certain circumstances, i.e. they cannot be explained in terms of linear superpositions of multiple adaptive single-task models. In order to specify alternative nonlinear control models, it is important to identify constraints on potential computational models of internal control structures that are implied by the experimental data. Nonlinear adaptive controllers form a challenging topic in Computational Neuroscience and are moreover highly relevant for the applications that are encompassed in the projects that involve the Otto Bock HealthCare company. Hand movements require complex controls that need to react adaptively to environmental influences. In particular, in a prosthetic device mutual adaptation processes must be realized between the remainders of neural motor-control and the adaptive intelligence of the prosthetic controllers. The Otto Bock hand prosthesis used in this project, receives input data derived from strongly variable high dimensional measurements of myoelectric activity. As a related problem, interpreting and enhancing strongly variable signals is studied in a muscular stimulation paradigm and forms a promising starting point for more ambitious neuroprosthetic projects. Finally, walking as a prime example of motor control combines local sensor-motor loops with computationally hard control problems. Otto Bock’s successful C-leg prosthesis will serve as a target application for adaptively controlled walking, where beyond presently existing technology myoelectrical inputs will be used and stabilization of double prostheses will be focussed on. In an initial phase the real prosthetic devices will be controlled by classical algorithms, whereas in a later stage more sophisticated control schemes and neuroprostheses using data recorded from the central nervous system will be tested for practical usability. The results of this section promise spin-offs also in other areas such as diagnosis of walking disorders, bipedal robotics, and exoskeletons.