Hints at attractor dynamics at work in premotor cortex during motor planning

STEFANO FERRAINA; BAGLIETTO, GABRIEL; DEL GIUDICE, PAOLO

Berlin

Conferencia; 11th FENS Forum of Neuroscience; 2018

We showed (Baglietto et al, Plos One 2017) that density-based clustering (DBC) provides efficient representations of multidimensional time series obtained from simultaneous electrophysiological recordings, and casts them in the form of symbolic sequences amenable to complexity estimates. Here we used DBC to analyze the spatio-temporal dynamics of premotor cortex during a countermanding reaching Multi-unit activity (MUA) was extracted from signals recorded using a 96-electrodes array. Performing DBC on the 96-dimensional MUA time series, we derived the corresponding discrete sequence of clusters' centroid.We first show that the joint analysis of such symbolic sequences for no-stop and stop trials confirms and extends previous observations from single-electrode analysis in (Mattia et al, J. Neurosci. 2013), providing predictive information on the outcome of a stop trial at the single-trial level.We then analyzed 1) the time course of the dispersion of MUA configurations within the clusters visited during the trial, 2) inter-trial variability of MUA configurations, 3) the velocity flow in the MUA configuration space, and find support to the hypothesis that the multi-dimensional neural dynamics during motor planning can be described as a sequence of transitions between attractors in the landscape defined by the density distribution in the MUA configuration space.