Temporal Scales of Cortical Interactions

Higher brain functions are attributed to the cortex. Over the years it became clear that information is encoded not only in the responses of individual neurons but also in the joint activity of populations of neurons. Based on theoretical studies it has been proposed that the temporally coordinated spiking activity of many neurons is a relevant variable for information processing (VON DER MALSBURG 1981, ABELES 1982b) and that cortical neurons organize dynamically into coherent functional groups (»cell assemblies«, HEBB 1949) that are distinguished by the coordinated activity of the participating neurons. Our research focuses on the development of analysis strategies for the identification of neuronal interactions and assembly activity. We attempt to decipher the spatial and temporal scales of dynamical neuronal interactions, and their relations to the external world (stimuli and/or behavior). In order to identify neuronal assemblies, simultaneously recorded neuronal spiking activity needs to be analyzed with respect to temporal structure. To that end we developed the »unitary event« analysis method (GRUN et al. 2002a, GRUN et al. 2002b) that detects the presence of conspicuous spike coincidences and evaluates their statistical significance. The analysis of simultaneously recorded neuronal activity in monkey primary motor and frontal cortex uncovered context-dependent, rapid changes in the patterns of coincident spike activity during performance of a delayed-pointing task (RIEHLE et al. 1997) or a delayed localization task (VAADIA et al. 1989, AERTSEN et al. 1991, VAADIA et al. 1991), respectively. Spike synchronization occurred accompanied by discharge rate modulations and in the absence of spike rate modulations depending on the details of the experimental protocol. The temporal precision of such synchronized events is in the range of a few ms (GRUN et al. 1999). Data suggest that the composition of significant coincidence patterns changes depending on the computational demands (GRUN et al. 2002b), which may be taken as an indication that different assemblies are activated in relation to behavior. In the unitary event analysis a number of different time scales have to be considered and affect different parameters of the signal. Here we specifically address the different temporal scales and give interpretations in respect to the dynamics of the neuronal processes.