Decision-related information at single-neuron resolution in human motor cortex and its implications for neuroprosthetics

点击翻译按钮获取中文摘要

Project Summary / Abstract The long-term goal of my research is to gain a holistic understanding of how movement commands are generated- including their relationship to sensory feedback and movement context- in order to restore movement to those who have lost it. When the ability to move is lost, due to spinal cord injury or disease, the ability to generate movement commands is still intact, but the command cannot reach the end effectors. One means of restoring function after such an injury is brain-computer interface (BCI) which records the movement commands directly from the brain and bypasses the injury to move external end effectors. These neural prostheses rely on accurate decoding of movement intention to perform the user’s desired action. While good control of these devices has been demonstrated, the control is not as quick as movement of a native limb. This may be due to oversimplification of how movement is decoded- in isolation from factors such as feedback or movement context. Cognitive processes that may be occurring at the same time are not accounted for when determining the user’s movement intention. The specific objectives of this proposal are to identify and characterize activity related to decision making in human motor cortex, at the single-cell level, and how it relates to movement command generation. Using participants enrolled in a clinical trial, we will record neural activity intracortically from motor cortex during decision-making tasks to identify decision-related activity and examine its relationship to movement intention. This activity has not been characterized in human cortex at this resolution, and the data collected in the proposed work will allow us to compare to models of decision making generated from non-human primate work and expand on the uniquely human ability to complete a variety of tasks in a single day. The proposed experiments will produce a valuable data set that will enable me to (1) identify how neurons represent both movement and decision-related activity, (2) capitalize on the decades of research on non-human primate decision making to identify appropriate models for decision-related activity in human motor cortex, (3) generate a more inclusive model by incorporating data collected during decision-related and movement generation tasks, using multiple end effectors and (4) integrate decision-related activity into BCI decoders to improve our ability to decode a BCI user’s movement intention and enhance the user’s control of the device (R00). This will provide insight - supplementing knowledge from non-human primate studies - as to how the brain decides to move.