Figure 1
From: The Energy Landscape of Neurophysiological Activity Implicit in Brain Network Structure
Conceptual Schematic. (A) A weighted structural brain network represents the number of white matter streamlines connecting brain regions. (B) While neurophysiological dynamics create rich time series of continuously-valued activity magnitudes, we study a simplified model in which each brain region is a binary object, being either active or inactive. (C) A schematic to provide an intuition regarding the nature of an energy landscape for the more general case of continuously-valued brain states. In our particular study, we simplify this picture by using a maximum entropy model to infer the landscape of predicted (binary) activity patterns – vectors indicating the regions that are active and the regions that are not active – as well as the energy of each pattern (or state). We use this mathematical framework to identify and study local minima in the energy landscape: states predicted to form the foundational repertoire of brain function.
