Neuroscience Cluster Scientific Retreat
Chicago Cultural Center, Chicago, IL
September 17, 2009



Aaron Suminski

Postdoctoral Scholar, Department of Organismal Biology & anatomy

Current brain-machine interfaces (BMI) largely rely on visual feedback to guide cursor or robot movements. It is well known, however, that patients suffering from the loss of proprioceptive sense exhibit considerable motor deficits such as slow and uncoordinated movements.  We designed an experiment to test the hypothesis that the addition of naturalistic proprioceptive feedback would improve the movement of a cortically-driven cursor.

 Two monkeys (Macaca mulatta) used a BMI[i] to move a visual cursor and hit a sequence of randomly placed targets while resting their arm in a two-link robotic exoskeleton.  The experiment consisted of three different conditions.  A micro-electrode array composed of 100 electrodes was used to record the spiking activity of an ensemble of neurons in primary motor cortex (MI) neurons under three different conditions.  In the first condition, Visual Feedback Decoding, the monkeys moved the cursor via the BMI voluntarily maintained a static arm posture in the robotic exoskeleton and thus did not receive veridical proprioceptive feedback regarding the cursor location.  Next, in the Visual and Proprioceptive Feedback Decoding condition, the monkeys controlled the cursor via the BMI while their arm was driven by the exoskeleton through the visual cursor trajectories, thereby providing the monkeys a veridical proprioceptive estimate of the cursor position.  Lastly, we included a control condition, Visual and Noisy Proprioceptive Feedback Decoding, in which the monkeys moved the cursor via the BMI while their arm was moved through a trajectory uncorrelated to the decoded trajectory of the visual cursor.

We demonstrate for the first time that proprioceptive feedback can be used together with vision to significantly improve control of a cursor driven by neural activity in the primary motor cortex. When the visual and proprioceptive feedback was congruent, the time to successfully complete the task decreased, and the cursor paths became straighter when compared with the incongruent feedback conditions.  These findings suggest that BMI control can be significantly improved in paralyzed patients with residual proprioceptive sense and provide the groundwork for augmenting cortically-controlled BMIs with multiple forms of natural or surrogate sensory feedback.

 

 

09/16/2009