Surface Electromyography for Brain-Machine Interface Applications
The project will focus on developing a wearable interface with spinal motor neurons. Non-invasive recordings of muscle activity can provide access to the neural activity of the innervating neural cells in the spinal cord. Therefore, they offer a window into the neural control strategies of movement by the central nervous system. As such, they can in principle be used for establishing a neural interface with humans. Moreover, since muscle electrical signals are easy to access and can be recorded non-invasively, this interface can be used both by patients (for rehabilitation technology) and by healthy individuals (for augmentation technology). This project will develop a wearable, high-density sensor for electromyographic recordings aimed at decoding the neural activity of spinal motor neurons during activities of daily living. The aim is to provide an additional control channel over external devices (e.g., smart watches, gaming, robots) that can be used in a robust way during daily activities and can overcome the limitations of other systems used for decoding the user intention (e.g., motion capture systems). The system will consist of several tens of electrodes embedded in textile to be worn at the wrist, similar to a tennis band, and embedded electronics for powering, conditioning, and processing of the signals. The signals will be decoded into the activity of motor neurons. The main challenge will be related to highly non-stationary conditions and brief contractions that will impose limitations on the statistical methods for blind separation of the neural cell activity. Moreover, methods for mapping the neural activity into control commands will be developed, with special emphasis on proportional and impedance control. Application scenarios will include virtual interfaces and gaming and will be tested over long-term conditions (several days of usage).