CHAOTIC DYNAMICS IN THE PRIMATE MOTOR CORTEX DEPEND ON MOTOR PREPARATION IN A DELAYED-POINTING TASK

This study investigates the influence of motor preparation on the neur al dynamics of time series generated by simultaneously recorded spike trains in the primary motor cortex (MI) of two monkeys performing a de layed-pointing task. A trial consisted of the following: (1) pre parat ory signal; (2) preparatory period lasting 600, 900, 1200, or 1500 ms; (3) response signal; and (4) pointing movement to a target light. We compared single and pairs of time series, with an overall count of mor e than 800 spikes, during the control period (i.e., before the onset o f the preparatory signal, n = 120 spike trains) and during the prepara tory period (n = 111), irrespective of its duration, to investigate th e existence of ''chaotic'' (i.e., seemingly unpredictable, yet determi nistic) dynamics in and across paired spike trains. Five neurons were characterized by the same dynamics in all recording periods out of a t otal of 20/120 and 14/111 deterministic dynamics observed during the c ontrol and preparatory periods, respectively. Interestingly, the value s of the geometric scaling properties of the neural dynamics during th e preparatory period were distributed over a broader range than during the control period. This suggests the disruption of a common or neutr al state of activity by the ongoing motor preparation. However, disreg arding the recording period, the majority (71%) of all paired time ser ies, including at least one ''chaotic'' spike train, showed determinis tic dynamics. Thus, the question is raised whether such dynamics bear some message, or whether this represents a ''side effect'' of some glo bal self-organizing dynamics within the cerebral cortex. Changes in th e temporal structure of MT spike trains induced by motor preparation s how that novel methods of analysis, based on the dynamical system appr oach, may help to reveal some principles underlying the organization o f brain cortical activity in different behavioral states.