Smooth pursuit eye movements in primates provide an accessible example of motor behavior guided by sensory inputs. Pursuit movements are controlled by cortical representations of target motion and access some of the same higher cortical regions implicated in planning and decision making. In preliminary work we have shown that pursuit behavior is variable but surprisingly precise, that the variability has a simple structure, and that it can be attributed mainly to errors in sensory estimates of target motion parameters. The surprisingly precise relationship between eye trajectories and target motion is established over time windows of 100 ms durations. Thus, pursuit gives us a remarkable situation - a genuine primate sensory-motor behavior in which the input-output relationship is computational simple, while relevant time scales are short enough that each cell can contribute at most a few spikes. Thus, in the equation in which behavior is a function of neural activity, both sides are much simpler than might have been expected. The potentially combinatorial complexity involved in a complete analysis of the neural code itself and the connection between spike trains and behavior is dramatically simplified. We propose 1) to understand the neural codes for sensory and motor signals at multiple levels of the neural circuit for pursuit, 2) to correlate the activity of single cortical, brainstem and cerebellar neurons with the trial-to-trial variability of motor output in awake, behaving animals, and 3) to bridge the gap from what we can measure (co-variation of neural and behavioral responses in single trials) to what we want to know (architecture and signal processing in the full sensory-motor circuit). The outcome of this line of research will be an understanding of how multiple cortical and sub-cortical areas work together to generate a single kind of voluntary movement. It will have direct impact on how we understand neurological disorders of movement, and on the consequences of disruptions or enhancements of correlations between neurons. Correlations and neuronal oscillations are an important feature of normal motor function, and this project will help us to understand how to understand their malfunctions and to design behavioral therapies to mitigate their disruption in epilepsy, nystagmus, and movement disorders.
Showing the most recent 10 out of 11 publications
