Abstract Nawrot FENS 2000

SPIKE COUNT VARIABILITY IN MOTOR CORTICAL NEURONS

M.P. Nawrot, A. Riehle (1), A. Aertsen, S. Rotter
Neurobiology and Biophysics, Biology III, University Freiburg, Germany
(1) CNRS-CRCN, Marseille, France.

Spiking activity of cortical neurons in vivo shows a considerable trial by trial variability, even when stimulus and behavioral conditions are identical for each repetition [1]. An empirical measure to quantify this variability is the Fano factor, defined as the ratio of variance and mean of spike counts across trials. In the visual system of alert mammals, this measure has consistently been found to yield values slightly above 1. Thus, it has been hypothesized that for cortical neurons the Fano factor generally takes values between 1 and 1.5 [2].
We analyzed single unit spike trains from 119 neurons in the monkey primary motor cortex, recorded during the performance of a delayed multi-directional hand-pointing task [3]. Our results show that Fano factors of motor cortical discharges are widely distributed, covering a range from 0.2 to 6, and in rare cases reach even beyond. Moreover, Fano factors may vary with the direction of the executed hand movement. More detailed analyses taking into account the time-resolved firing rate profiles [4] revealed that many of the recorded neurons exhibit systematic temporal changes of trial by trial variability throughout the trial.
Taken together, our findings demonstrate that the spiking process of primary motor cortical neurons cannot be described by a rate-modulated Poisson process. Instead, they indicate that a more elaborate point process model, possibly with time-dependent statistics, is required to adequately describe motor cortical unit activity.