An EMG-Based Biomimetic Prosthetic Hand for Grasping

The development ot interlaces that link the human musculoskclctal system with robotic devices has been a major area of research within rehabilitation robotics. Most of the research is focused on restoration of motor and sensory functions to those with de­ generative diseases, injury or amputees. The basic goal is to enhance capability for independent living and vocational productivity by restoring the physical functionality through use of prosthesis. One of the key focus has been upper limb prostheses for people with manipulative disabilities. Bio-signals driven prosthetic hands have been found to be suitable; wherein control is through conveying human 's intention to the prosthesis. There arc two possible bio-signal based schemes covering the approaches for conveying human's intention to the prosthesis. ¿ Electroencephalogram (EEG) based approaches ¿ Elcctromyogram (EMG) based approaches EEG based approaches arc implemented through an interface between the brain and the prosthetic hand to be controlled. The activity of the brain is recognized based on the EEG signals (5). On successful recognition of brain's activity; the prosthesis emulate the amputee's intention through the interface. Due to localization of brain activities and multidimensional aspect of the EEG signals, analysis and classificat10n of EEG signals arc challenging. Moreover, the appropriate number of channels as well as their specific location on the scalp requires identification. Failing to do so results in degradation of system performance. In many cases, there is no clear agreement about the number and location of necessary channels to collect EEG signals.