Transcranial ultrasound stimulation effect in the redundant and synergistic networks consistent across macaques

Marilyn Gatica, Cyril Atkinson-Clément, Pedro A. M. Mediano, Mohammad Alkhawashki, James Ross, Jérôme Sallet, Marcus Kaiser

Abstract

Low-intensity transcranial ultrasound stimulation (TUS) is a non-invasive technique that safely alters neural activity, reaching deep brain areas with good spatial accuracy. We investigated the effects of TUS at the level of macaque using a recent metric, the synergy minus redundancy rank gradient, that quantifies different kinds of causal neural information processing. We analyzed this high-order quantity on the fMRI data after TUS in two targets: the supplementary motor area (SMA-TUS) and the frontal polar cortex (FPC-TUS). The TUS produced specific changes at the limbic network at FPC-TUS and the motor network at SMA-TUS and altered, in both targets, the sensorimotor, temporal, and frontal networks, consistent across macaques. Moreover, there was a reduction in the structural and functional coupling after both stimulations. Finally, the TUS changed the intrinsic high-order network topology, decreasing the modular organization of the redundancy at SMA-TUS and increasing the synergistic integration at FPC-TUS.

Introduction

While non-invasive brain stimulation techniques like transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS) have become staples in neuroscience, they often lack the depth and precision required to target subcortical structures without affecting overlying cortex. Low-intensity transcranial ultrasound stimulation (TUS) has emerged as a powerful alternative, capable of reaching deep-seated brain regions with millimeter-level spatial accuracy.

However, the mechanism by which TUS reorganizes large-scale neural communication remains poorly understood. Most existing research focuses on pairwise functional connectivity—simple correlations between two regions. This study moves beyond pairwise interactions to explore how TUS affects the "high-order" informational architecture of the brain, specifically focusing on the balance between redundant and synergistic information processing.

Methods

The researchers analyzed resting-state fMRI data from three macaques across three conditions: a control (non-TUS) state, stimulation of the Supplementary Motor Area (SMA), and stimulation of the Frontal Polar Cortex (FPC). TUS was applied in 40-second bursts.

To quantify the complex interactions between multiple brain regions, the team utilized Integrated Information Decomposition ($\\Phi ID$). This framework allows for the separation of neural information into two primary components:

  1. Redundancy: Information that is shared across multiple neural sources, providing robustness and reliability.
  2. Synergy: Emergent information that arises only when multiple parts of the system work together, representing complex integration.
Conceptual illustration of redundancy and synergy in signal processing.

By calculating a "synergy minus redundancy rank gradient," the researchers could map which brain regions act as redundant hubs versus synergistic integrators. These informational maps were then compared against the brain's physical "wiring" (structural connectivity) derived from diffusion MRI to assess structure-function coupling.

Key Results

Target-Specific Network Reorganization

The effects of TUS were not localized solely to the stimulation site but radiated through specific functional networks. Stimulating the FPC, a region associated with high-level cognition, primarily shifted the limbic network (including the striatum and hippocampus) toward more redundant interactions. Conversely, stimulating the SMA, a motor-related area, shifted the motor network (including the pallidum and posterior thalamus) toward redundancy, while pushing frontal areas like the ventrolateral prefrontal cortex toward synergy.

The Structure-Function "Divorce"

A significant finding was the decoupling of functional information flow from the underlying structural architecture. In the control state, there is a moderate correlation between the brain's physical wiring and its redundant information flow (approx. $\\rho = 0.24$). Following TUS, this correlation dropped significantly (to approx. $\\rho = 0.19$). This suggests that TUS temporarily "frees" the functional network from its structural constraints, allowing for a flexible reorganization of neural communication.

Altered Information Topology

The study applied graph theory to examine how information is organized across the whole brain. Redundant information is typically highly modular, meaning it is segregated into distinct, reliable pockets. TUS disrupted this organization: SMA-TUS decreased the modularity of redundancy, effectively "blurring" the boundaries of shared information in the motor system. Meanwhile, FPC-TUS increased the global integration of synergy, enhancing the brain's ability to process complex, emergent information across disparate regions.

Discussion and Conclusion

This study demonstrates that TUS is a potent tool for modulating the high-order informational dynamics of the brain. The consistency of these results across all three macaques suggests that the shift toward redundancy in targeted networks and the overall decoupling from structural constraints are fundamental biological responses to ultrasound.

While the sample size is small and the use of anesthesia may influence connectivity patterns, these findings provide a roadmap for future clinical applications. By understanding how TUS rebalances synergy and redundancy, researchers may eventually use this technology to "re-tune" information flow in human patients suffering from neurological or psychiatric disorders characterized by disrupted neural integration.

Summary updated

Abstract

Abstract Low-intensity transcranial ultrasound stimulation (TUS) is a non-invasive technique that safely alters neural activity, reaching deep brain areas with good spatial accuracy. We investigated the effects of TUS at the level of macaque using a recent metric, the synergy minus redundancy rank gradient, that quantifies different kinds of causal neural information processing. We analyzed this high-order quantity on the fMRI data after TUS in two targets: the supplementary motor area (SMA-TUS) and the frontal polar cortex (FPC-TUS). The TUS produced specific changes at the limbic network at FPC-TUS and the motor network at SMA-TUS and altered, in both targets, the sensorimotor, temporal, and frontal networks, consistent across macaques. Moreover, there was a reduction in the structural and functional coupling after both stimulations. Finally, the TUS changed the intrinsic high-order network topology, decreasing the modular organization of the redundancy at SMA-TUS and increasing the synergistic integration at FPC-TUS.
BibTeX
@inproceedings{gatica2024transcranial,
  author = {Marilyn Gatica and Cyril Atkinson-Clément and Pedro A. M. Mediano and Mohammad Alkhawashki and James Ross and Jérôme Sallet and Marcus Kaiser},
  title = {{Transcranial ultrasound stimulation effect in the redundant and synergistic networks consistent across macaques}},
  booktitle = {Network},
  year = {2024},
  doi = {10.1162/netn\_a\_00388},
}
Figure