UIUC Quant Brownbag
Theories of minds: Connecting cognitive architecture models to brain imaging data
Catherine Sibert - University of Groningen
Cognitive architecture models can provide a high-level, whole-brain account of
cognition, but have not seen wide use in the neuroscience research community due
to a perceived lack of connection between the theoretical model elements and the
biological reality of the brain. Recent work has tried to bridge this gap by using the
structural configuration of architecture models, specifically the Common Model of
Cognition (CMC), as a framework for Dynamic Causal Modeling (DCM) analyses of
large scale fMRI data, both during tasks and at rest. The high-level framework
proposed by the CMC produces higher quality predictions of human brain data than
several plausible alternatives, suggesting both the presence of a consistent
underlying structure of cognition in the brain as well as a sense of its shape and
dynamics. These results demonstrate how the analysis of brain signaling data can
benefit from the addition of cognitive modeling theories, as well as showing how
brain data can be used to validate, refine, and improve cognitive architecture
frameworks. Combining these tools and approaches provides many challenges and
opportunities for gaining a deeper understanding of the structure and mechanisms of
cognition.