Parcels and particles: Markov blankets in the brain.

Karl J Friston,Tahereh S Zarghami,Loïc Magrou,Erik D Fagerholm,Inês Hipólito,Thomas Parr,Adeel Razi

doi:10.1162/netn_a_00175

Abstract

At the inception of human brain mapping, two principles of functional anatomy underwrote most conceptions—and analyses—of distributed brain responses: namely, functional segregation and integration. There are currently two main approaches to characterizing functional integration. The first is a mechanistic modeling of connectomics in terms of directed effective connectivity that mediates neuronal message passing and dynamics on neuronal circuits. The second phenomenological approach usually characterizes undirected functional connectivity (i.e., measurable correlations), in terms of intrinsic brain networks, self-organized criticality, dynamical instability, and so on. This paper describes a treatment of effective connectivity that speaks to the emergence of intrinsic brain networks and critical dynamics. It is predicated on the notion of Markov blankets that play a fundamental role in the self-organization of far from equilibrium systems. Using the apparatus of the renormalization group, we show that much of the phenomenology found in network neuroscience is an emergent property of a particular partition of neuronal states, over progressively coarser scales. As such, it offers a way of linking dynamics on directed graphs to the phenomenology of intrinsic brain networks.

Highlights

A persistent theme in systems neuroscience, especially neuroimaging, is the search for principles that underlie the functional anatomy of distributed neuronal processes
In recent thinking about functional integration, people have turned to formal accounts of processing in the brain (e.g., Bastos et al, 2012; Keller & Mrsic-Flogel, 2018; Parr & Friston, 2018; Rao & Ballard, 1999; Spratling, 2008) to understand the nature of message passing on graphs, where edges correspond to connectivity and nodes correspond to neuronal populations
We review the notion of Markov blankets and how recursive applications of a partition or parcellation of states into Markov blankets allows one to express dynamics at increasing scales

Summary

INTRODUCTION

A persistent theme in systems neuroscience, especially neuroimaging, is the search for principles that underlie the functional anatomy of distributed neuronal processes. Following Bayesian model reduction, we have a sparse Jacobian or directed, weighted adjacency matrix describing the dynamical coupling between univariate states of 1,024 particles (see Figure 4) Note that it would have been possible to reevaluate the Jacobian using another dynamic causal model of the eigenstates at any particular level and use Bayesian model reduction to eliminate redundant coupling parameters This is an interesting alternative to using the estimates of the Jacobian based upon the first-order approximation at the smallest scale. This means that I should be able to tell you whether you have “seen something” in the past minute or so by examining your brain activity at this moment in time

CONCLUSION

Findings

FUNDING INFORMATION