Chaotic Neuronal Oscillations in Spontaneous Cortical-Subcortical Networks

TL;DR

This study reveals that various spontaneous neural oscillations across brain regions are genuine chaotic systems, which can be modeled and predicted using chaotic attractors, suggesting deterministic processes underlying brain rhythms.

Contribution

It demonstrates that multiple brain oscillations are chaotic and can be reconstructed as attractors, providing new insights into their deterministic nature and predictability.

Findings

Neural oscillations are genuine chaotic time series.

Reconstructed attractors enable high-precision short-term prediction.

Oscillations across regions can be approximated by the same function despite phase differences.

Abstract

Oscillatory activities are widely observed in specific frequency bands of recorded field potentials in different brain regions, and play critical roles in processing neural information. Understanding the structure of these oscillatory activities is essential for understanding the brain function. So far many details remain elusive about their rhythmic structures and how these oscillations are generated. We show that many oscillatory activities in spontaneous cortical-subcortical networks, such as delta, spindle, gamma, high-gamma and sharp wave ripple bands in different brain regions, are genuine chaotic time series which can be reconstructed as chaotic attractors through appropriately selected embedding delay and dimension. The reconstructed attractors are approximated by a simple radial basis function enabling high precision short-term prediction. Simultaneously recorded oscillatory activities in multiple brain regions differ greatly in term of temporal phase and amplitude but can be approximated by the same function. Our results suggest that neural oscillations are produced by deterministic chaotic systems. The occurrence of neural oscillation events is predetermined, and the brain possibly knows when and where the information will be processed and transferred in the future time as a result of the deterministic dynamic.