Human Brain’s Functional Connectivity Boils Down to Three Patterns in Time and Space

Human brain function is predominantly studied using noninvasive, resting-state, functional magnetic resonance imaging (fMRI). However, the application of this technology since the 1990s has led to seemingly contradictory concepts regarding the large-scale functional organization of the human brain.

A collaborative team including scientists from Emory University, the University of California, Los Angeles (UCLA), the Montreal Neurological Institute, Vanderbilt University, and the National University of Singapore, have now bridged the gap between two broad categories of large-scale brain organization and showed the most functional phenomena in the human brain can be reduced to combinations of three spatiotemporal patterns. These patterns explain most global spatial structures of the brain responsible for functional connectivity and unify observations made using resting-state functional MRI that were previously thought to be different.

Lucina Uddin, PhD, professor of psychiatry and biobehavioral sciences and director of the Brain Connectivity and Cognition Laboratory at the UCLA Semel Institute for Neuroscience and Human Behavior, is an author of the study.

“Functional magnetic resonance imaging suggests the brain has a globally coherent spatial structure, but there is not yet consensus among scientists on the proper way to catalog this structure. We show that a small number of spatiotemporal patterns can do the job,” said Lucina Uddin, PhD, professor of psychiatry and biobehavioral sciences and director of the Brain Connectivity and Cognition Laboratory at the UCLA Semel Institute for Neuroscience and Human Behavior.

The findings were published in an article titled, “A parsimonious description of global functional brain organization in three spatiotemporal patterns,” in the journal Nature Neuroscience.

“We showed that a wide range of previously observed empirical phenomena are manifestations of three main spatiotemporal patterns,” said Taylor Bolt, PhD, a statistician at the UCLA Semel Institute and the first author of the paper.

MRI measures are based on spontaneous low-frequency fluctuations that are blood oxygenation level dependent (BOLD). These fluctuating signals reflect synchronous neural activity between different regions of the brain and are organized into global patterns that span functional systems of cognition, perception, and action. Spontaneous fluctuations have been subjected to increasingly complex analytical techniques, leading to competing descriptions of large-scale functional brain organizations.

These global patterns are characterized into two broad categories: zero-lag and time lag synchrony. Zero-lag synchrony represents “stationary waves” where an instantaneous dependence exists between two signaling time courses or there is no time lag between two BOLD signals. This can be visualized as simultaneous synchrony of brain regions across the cortex. Time-lag synchrony, on the other hand, represents “traveling waves” where the dependence between two signaling time courses is separated by a delay.

To extract traveling wave patterns, the authors applied PCA (principal component analysis) to complex BOLD signals, that were derived from the original using a mathematical transformation technique.

There has been “little attempt to synthesize findings across different approaches,” the researchers said. Uddin compared this lack of a unified concept to the Indian parable where blindfolded men encountering different parts of an elephant arrive at different descriptions of the same animal.

“We sought to unify observed phenomena across these two categories in the form of three low-frequency spatiotemporal patterns composed of a mixture of standing and traveling wave dynamics,” the authors noted.

The network structure of functional connectivity in human brains is explained by three spatiotemporal patterns. [Bolt T., Nat. Neuro., 2022]

The authors hypothesized that representations of stationary and traveling waves capture different aspects of a small number of spatiotemporal patterns. In testing this hypothesis, the authors found that several earlier observations, including functional connectivity gradients, task-positive/task-negative anti-correlation patterns, time-lag propagation patterns, and the functional connectome network structure, could be unified in a framework modeling both standing and traveling wave structures.

These results present a global functional organization of the brain that could inspire investigations on coordinating principles of brain activity.

Leave a Reply

%d bloggers like this: