PROJECT SUMMARY Multiple neurological diseases [e.g. spinal cord injury (SCI), amyotrophic lateral sclerosis (ALS), brain stem stroke] can all result in severe and devastating limb paralysis. A recent comprehensive assessment found that >200,000 patients suffer from tetraplegia or severe tetraparesis that prevents completion of basic activities of daily living that require arm and hand functions. Surveys of such patients have indicated that improvement of arm and hand function is a top priority. There are no current therapies or assistive devices that can aid patients with tetraplegia or severe tetraparesis to experience restoration of reaching and grasping functionality. Our proposal aims to test methods to enable such patients to directly control a complex robotic arm and hand with the capacity to perform a set of clinically relevant tasks. Our specific goals are to leverage the stability of ECoG to establish robust robotic control that is stable across a period of at least 8 weeks without need for recalibration. Our published data along with new preliminary data supports the notion that ECoG signals can allow a paralyzed individual to learn complex neuroprosthetic control that requires no additional training. We will compare two decoding methods and their ability to enable long-term stable ‘plug-and-play’ complex control. We then aim to further boost robustness of real-world control in two ways. First, we will track fluctuations in neural states to reduce decoding errors; this is key for long-term continuous accurate control. Second, we will test a system that can assist with pre-shaping the robot during neuroprosthetic control. Together, our aims will determine the feasibility of complex control of neuroprosthetic technology in a target population of paralyzed patients with severe disability. We will determine how well ECoG can enable stable and intuitive control of a robotic arm and hand that can enable reaching, grasping and flexible manipulation of objects. We strongly believe that demonstration of these outcomes will drive the field towards clinically viable neuroprosthetic control and thereby dramatically improve the quality of life for paralyzed patients.

PROJECT NARRATIVE There are >200,000 patients living with severe motor disability as a result of tetraplegia or severe tetrapereisis resulting from multiple neurological disorders. Brain-Computer Interfaces (BCIs) can reduce the extent of disability by allowing direct neural control of assistive devices. Our proposal aims to test whether a BCI using electrocorticography can allow such patients to control a complex robotic arm to perform a range of clinically relevant dexterous tasks.