Information processing in large-scale cerebral networks: the causal connectivity approach

Today, cognitive functions are considered to be the offspring of the activi ty of large-scale networks of functionally interconnected cerebral regions. The interpretation of cerebral activation data provided by functional imag ing has therefore recently moved to the search for the effective connectivi ty of activated regions, which aims at understanding the role of anatomical links in the activation propagation. Our assumption is that only causal co nnectivity can offer a real understanding of the links between brain and mi nd. Causal connectivity is based on the anatomical connection pattern, the information processing within cerebral regions and the causal influences th at connected regions exert on each other. In our approach, the information processing within a region is implemented by a causal network of functional primitives, which are the interpretation of integrated biological properti es. Our choice of a qualitative representation of information reflects the fact that cerebral activation data are only the approximate view, provided by imaging techniques, of the real cerebral activity. This explicit modelin g approach allows the formulation and the simulation of functional and phys iological assumptions about activation data. Two alternative models explain ing results of the striate cortex activation described by Fox and Raichle ( Fox PT, Raichle ME (1984) J. Neurophysiol 51:1109-1120; Fox PT, Raichle ME (1985) Ann Neurol 17:303-305) are provided as an example of our approach.