Upper limb reaching and grasping movements require complex cortical control circuits involving both motor- control outputs and real-time somatosensory feedback. Neurological disorders such as strokes, brain trauma, and spinal cord injury may result in a loss of the ability to perform these tasks. Many teams, including our own, are working to restore upper extremity function by using human neural signals to control the movements of a robotic limb with multiple degrees of freedom [1-3]. However, without somatosensory feedback, even the most basic limb movements are difficult to perform in a fluid and natural manner [4, 5]. There have only been a limited number of human studies exploring how to generate somatosensory feedback. Using subdural electrocorticography (ECoG) grids placed on the human primary somatosensory (S1) hand area in patients with epilepsy who require intracranial monitoring, we propose studies directed toward understanding how somatosensation is cortically encoded and how we can restore upper extremity somatosensation via electrical stimulation. To accomplish this, I have assembled a multidisciplinary mentoring team, led by Dr. Gianluca Lazzi, with an established history of success in mentoring early investigators. From my mentoring team, I plan on learning about neural modeling, study design and biostatistics, and medical device development. My long-term goal is to become an independent NIH-funded neurosurgeon-scientist who makes significant contributions to our understanding of sensorimotor integration.
In Aim 1 we will use the participants own ECoG responses to real touch to guide a systematic mapping of stimulation parameter space to find distinct percepts of somatosensation. Much like how clinical neurostimulators such as deep brain stimulators (DBS) for movement disorders and vagus nerve stimulators (VNS) are therapeutic only at specific stimulation settings, we hypothesize that we will find specific stimulation combinations that result in different types of somatosensation.
In Aim 2 we will compare task performance using artificial somatosensation versus native touch.
In Aim 3 we will quantify how real touch and artificial somatosensation generated by ECoG stimulation differ in response time between real touch/stimulation and participant perception. These results and the mentoring provided through this K23 program will be a critical foundation for my transition to an independent investigator in sensorimotor integration.
For patients with upper extremity dysfunction caused by strokes, brain trauma, or spinal cord injury, brain- machine interface (BMI) technologies may one day restore function by utilizing cortical neural signals to control the movements of a robotic limb with multiple degrees of freedom. However, the current generation of BMI devices are guided by visual feedback alone, without somatosensory feedback. Our proposal is designed to increase our understanding of sensorimotor integration by examining how varying stimulation parameters affect perceived somatosensation (Aim 1), examine task performance using artificial somatosensation versus native touch (Aim 2), and compare real touch to artificial somatosensation for response time (Aim 3).
