Dynamic Synchrony of Local Cell Assembly

Abstract

In the present paper, we focus on the problem of the dynamic size of a cell assembly and discuss how we can detect synchronized firing of a local cell assembly consisting of closely neighboring neurons in the working brain. A local cell assembly is difficult to detect because of the problem of spike overlapping of neighboring neurons, which cannot be overcome by ordinary spike-sorting techniques. We introduce a unique technique of spike-sorting that combines independent component analysis (ICA) and an ordinary sorting method to separate individual neighboring neurons and analyze their firing synchrony in behaving animals. One of our experiments employing this method showed that some closely neighboring neurons in the monkey prefrontal cortex have dynamic and sharp synchrony of firing reflecting local cell assemblies during working-memory processes. Another experiment showed that our other method (ICSort) of novel spike-sorting by ICA using special electrodes (dodecatrodes) can distinguish firing signals from the soma and those from the dendrites of individual neurons in behaving rats and suggests that the somatic and dendritic signals have different roles in information processing. This indicates that functional connectivity among neurons may be more dynamic and complex and spikes from the soma and dendrites of individual neurons should be considered in the investigation of the activity of local cell assemblies. We finally propose that detailed and real features of a local cell assembly consisting of closely neighboring neurons should be examined further and detection of local cell assemblies could be applied to the development of neuronal prosthetic devices, that is, brain-machine interfaces (BMIs).