The brain is assumed as a Distributed Intelligent Processing System (DIPS) composed of agents having some specialization in solving defined problems, such that reasoning becomes a cooperative activity among as much as possible decentralized and loosely coupled collection of these agents that may provide the solution of a given problem. DIPS’ intelligence emerges as a function of how versatile are the relations shared by these agents; of how plastic may be the commitments for actions among them, and of agent specialization. The complexity of agent enrollment (h(c)) in solving a task is proposed to be quantified by the entropy of the available communication resources minus the entropy of the communication resources actually used to solve the task. The linear correlation coefficient calculated for the task event related activity was as taken as a measurement of the possibility of information exchange among neurons from different cortical areas, and it was used to calculated h(c). The calculated h(c) values for the EEG activity recorded by each 10/20 system electrode while the individual is solving a cognitive task are collor encoded to produce the corresponding Cognitive Brain Mapping (CBM). CBMs are different for different cognitive tasks and also varied with the different experimental. The calculated h(c) values for different cognitive tasks correlate with individual IQ measured by means of standard tests. Also, response times and error rates were found to be correlated with h(c).



ROCHA, A. F. et al. (2011) The Brain as Distributed Intelligent Processing Systes: An EEG study. PloS One, 6: e17355