HIDDEN MARKOV MODELING OF SIMULTANEOUSLY RECORDED CELLS IN THE ASSOCIATIVE CORTEX OF BEHAVING MONKEYS

A widely held idea regarding information processing in the brain is th e cell-assembly hypothesis suggested by Hebb in 1949. According to thi s hypothesis, the basic unit of information processing in the blain is an assembly of cells, which can act briefly as a closed system, in re sponse to a specific stimulus. This work presents a novel method of ch aracterizing this supposed activity using a hidden Markov model. This model is able to reveal some of the underlying cortical network activi ty of behavioural processes. In our study the process in hand was the simultaneous activity of several cells recorded from the frontal corte x of behaving monkeys. Using such a model we were able to identify the behavioural mode of the animal and directly identify the correspondin g collective network activity. Furthermore, the segmentation of the da ta into the discrete states also provides direct evidence for the stat e dependence of the short-time correlation functions between the same pair of cells. Thus, this cross-correlation depends on the network sta te of activity and not on local connectivity alone.