[Neurocognition and PET. Strategies for data analysis in activation studies on working memory]

In cognitive neuroscience regional cerebral blood flow (rCBF) imaging with positron-emission-tomography (PET) is a powerful tool to characterize different aspects of cognitive processes by using different data analysis approaches. By use of an n-back verbal working memory task (varied from 0- to 3-back) we present cognitive subtraction analysis as basic strategy as well as parametric and covariance analyses and discuss the results.

Correlation analyses were performed using the individual performance rate as an external covariate, computing inter-regional correlations, and as network analysis applying structural equation modelling to evaluate the effective connectivity between the involved brain regions.

Subtraction analyses revealed a fronto-parietal neuronal network also including the anterior cingulate cortex and the cerebellum. With higher memory load the parametric analysis evidenced linear rCBF increases in prefrontal, pre-motor and inferior parietal areas including the precuneus as well as in the anterior cingulate cortex. The rCBF correlation with the individual performance as external covariate depicted negative correlations in bilateral prefrontal and inferior parietal regions, in the precuneus and the anterior cingulate cortex. The network analysis demonstrated mainly occipito-frontally directed interactions which were predominantly left-hemispheric. Additionally, strong linkages were found between extrastriate and parietal regions as well as within the parietal cortex.

The data analysis approaches presented here contribute to an extended and more elaborated understanding of cognitive processes and their different sub-aspects.