Presently, those suffering from tetraplegia after an injury or illness have little hope for significant motor restoration. Brain-computer interfaces (BCIs) have shown tremendous promise for controlling robotic arms and computer tablets as well as restoring somatosensory function. Dr. Willsey’s work focuses on developing an entry level brain-computer interface to help restore motor function for people with paralysis.
The seminar covers the basic principles of BCI technology, available recording systems (e.g., electroencephalography, electrocorticography, and microelectrode arrays), and controllable effectors, along with development of a state-of-the-art decoder. The talk will also describe a state-of-the-art decoder, used to translate neural activity into a control signal for external devices. Specifically, this temporally convolved neural network decoder, which incorporated a time series of neural data into the decoding algorithm, allowed the simultaneous movement of 3 finger groups, of which the thumb moves in 2D, and acquired 76 finger targets per minute. Furthermore, a human participant with tetraplegia controlled a virtual quadcopter navigating a ringed obstacle course in a physics-based simulation using these decoded finger movements. Finally, Dr. Willsey will review several new companies that are commercializing BCI technology in early clinical trials and discuss the current challenges in translating BCIs for eventual human use.
