Computational Pipeline to Infer Micro-, Meso-, and Macro-Scale Brain-State Transitions

Attractors are a framework for understanding and modeling the neural dynamics argued to underlie diverse cognitive phenomena, be it working memory, learning, neural oscillations, or arousal. Neural attractors have been argued to be present at the scale of single-cells, networks of cells, and between systems-scale regions of the whole-brain. Here, we describe a methodology inspired by statistical mechanics to calculate low-dimensional effective energy landscapes from neural data providing insight into the probability of brain-state transitions, which allow us to infer the brain's approximate attractor structure.