Society For Neuroscience October 2005, Poster Presentation

M. Versace; S. Grossberg

Boston University, Boston, MA, USA

A model of cortico-cortical and thalamo-cortical learning and information processing is presented that investigates the functional significance of higher-order specific thalamic nuclei as well as nonspecific thalamic nuclei, such as the midline and intralaminar nuclei. The model proposes how synchronization of neuronal spiking occurs within and across multiple brain regions and thalamic nuclei and simulates the functional link between synchronization and synaptic plasticity. The model is described in terms of spiking neurons that obey Hodgkin-Huxley type dynamics. It explicates within this setting various operations from Adaptive Resonance Theory to control information processing and learning, notably the control of matching processes that can lead either to resonance and learning, or to reset and memory search. Data about cortical and thalamic neurophysiology and anatomy and single cell biophysics are linked to functional properties such as resonance/learning, reset/search, top-down attention, and synchrony. In particular, the model simulates data on single cell biophysics, both cortical (Williams and Stuart, 1999; Saar et al, 2001; Larkum and Zhu, 2002) and subcortical (Sillito et al, 1994; Murphy et al, 1999; Jones, 2002; Crabtree and Isaac, 2002), and aggregate cell recordings (current-source densities and local field potentials; Kraus et al 1994; Heynen and Bear, 2001). It functionally links single cell and large-scale oscillations, notably in the gamma frequency band (Singer, 1990; Basar et al, 1991; Engel et al, 1991; Singer and Gray, 1995; Steriade and Amizca, 1996; Pantev, 1995; Miltner et al, 1999; Friedman-Hill et al, 2000), and clarifies the functional meaning of different oscillation frequencies.