Superlinear population encoding of dynamic hand trajectory in primary motor cortex

Liam Paninski, Shy Shoham, Matthew Fellows, Nicholas Hatsopoulos, and John Donoghue

Appeared as
Journal of Neuroscience 24: 8551-8561.

Neural activity in primary motor cortex (MI) is known to correlate with hand position and velocity. Previous descriptions of this tuning have a) been linear in position or velocity, b) depended only instantaneously on these signals, and/or c) not incorporated the effects of interneuronal dependencies on firing rate. We show here that many MI cells encode a nonlinear function of the full time-varying hand trajectory. Roughly 20% of MI cells carry information in the hand trajectory beyond just the position, velocity, and acceleration at a single time lag. Moreover, roughly one third of MI cells encode the trajectory in a significantly superlinear fashion; as one consequence, even small position changes can dramatically modulate the gain of MI cells' velocity tuning, in agreement with recent psychophysical evidence (Hwang et al. '03). We introduce a compact nonlinear filter model that predicts the complex structure of the spatiotemporal tuning functions described in earlier work (Paninski et al. '04). Finally, observing the activity of neighboring cells in the MI network significantly increases the predictability of a single MI cell's firing rate; however, we find interneuronal dependencies in MI to be much more locked to external kinematic parameters than those recently described in the hippocampus (Harris et al. '03). Nevertheless, this neighbor activity is approximately as informative as the hand velocity, supporting the view that neural encoding in MI is best understood at a population level.
Reprint (pdf, 540K)  |  Related work on spike-triggered analysis techniques

Related work on neural coding in motor cortex  |  Liam Paninski's home