Complexities of Cortex and the Need for Detailed Models

Abstract

In the next chapters, we shall consider mammalian cortical structures, including the neocortex, piriform (olfactory) cortex, entorhinal cortex, and hippocampus: their neurons, synapses and gap junctions, and some collective network phenomena that involve hundreds to thousands of cells. The latter phenomena include oscillations, synchronized firings, and epileptic-like activities: phenomena which are truly collective but where insight is possible even if one does not account exactly for what every neuron is doing at every time. We shall also consider cell assemblies, where the identity of the neurons participating in an assembly is presumed to matter but where a network model does not draw one-to-one correspondences between model neurons and particular experimentally recorded neurons. The model seeks to predict properties of the assemblies, not their exact contents. In a word, our approach has some resemblance to the study of CPGs – we take into account electrophysiological and synaptic details – yet is also different, as there are too many neurons now, the neurons are not identifiable, and we seek to explain more generic properties of the system rather than draw exact model-experiment neuron-neuron correlations. But not too generic.