Interpreting Neural Population Activity During Feedback Motor Control

2015-05-01T00:00:00Z (GMT) by Matthew Golub
The motor system routinely generates a multitude of fast, accurate, and elegant movements.
In large part, this capacity is enabled by closed-loop feedback control systems in the brain.
Brain-machine interfaces (BMIs), which translate neural activity into control signals for driving
prosthetic devices, also engage the brain’s feedback control systems and offer a promising
experimental paradigm for studying the neural basis of feedback motor control. Here, we address
both the engineering challenges facing current BMI systems and the basic science opportunities
afforded by them.
Previous studies have demonstrated reliable control of the direction of movement in cursorbased
BMI systems. However, control of movement speed has been notably deficient. We provide
an explanation for these observed difficulties based on neurophysiological studies of arm
reaching. These findings inspired our design of a novel BMI decoding algorithm, the speeddampening
Kalman filter (SDKF) that automatically slows the cursor upon detecting changes
in decoded movement direction. SDKF improved success rates by a factor of 1.7 relative to a
standard Kalman filter in a closed-loop BMI task requiring stable stops at targets.
Next, we transition toward leveraging the BMI paradigm for basic scientific studies of feedback
motor control. It is widely believed that the brain employs internal models to describe our
prior beliefs about how an effector responds to motor commands. We developed a statistical
framework for extracting a subject’s internal model from neural population activity. We discovered
that a mismatch between the actual BMI and the subjects internal model of the BMI explains
roughly 65% of movement errors. We also show that this internal model mismatch limits movement
speed dynamic range and may contribute toward the aforementioned known difficulties in
control of BMI movement speed.