A correlation method was developed to examine functional interactions between brain regions, by correlating either regional cerebral metabolic rates for glucose or regional cerebral blood flows, as determined by positron emission tomography (PET) in humans. There was a loss of frontal-parietal functional associations with age, and a further reduction in patients with dementia of the Alzheimer type (DAT), suggesting fewer corticocortical connections. In patients with well controlled hypertension, correlations involving regions in the vascular watershed areas were significantly reduced. In humans in whom regional cerebral blood flow (rCBF) was measured with PET during two visual processing tasks, correlations among visual brain areas were significant in the right but not the left hemisphere, suggesting a more important role for the right hemisphere. A systems-level neural network-model, fitted to rCBF PET data, permitted determination of the brain regions and their interactions that were involved in two visual processing tasks. A computer simulation model was developed to to explore the neurobiological substrates of observed correlational patterns. A multiple regression/discriminant analysis involving PET regional interdependencies distinguished DAT patients from controls. In young adults with Down syndrome, PET values in language areas could be used in a discriminant function to distinguish their PET scans from those of controls. A multiple regression/discriminant analysis applied to PET scans of patients with Obsessive-Compulsive Disorder obtained before and after pharmacotherapy allowed discrimination between patients who responded to drug from those who did not.
