Antero-posterior versus lateral vestibular input processing in human visual cortex

Felipe Aedo-Jury, Benoit R. Cottereau, Simona Celebrini, Alexandra Séverac Cauquil

Abstract

Visuo-vestibular integration is crucial for locomotion, yet cortical mechanisms involved remain poorly understood. We combined binaural monopolar galvanic vestibular stimulation (GVS) and functional magnetic resonance imaging (fMRI) to characterize the cortical networks activated during antero-posterior and lateral stimulations in humans. We focused on functional areas that selectively respond to egomotion-consistent optic flow patterns: the human middle temporal complex (hMT+), V6, the ventral intraparietal (VIP) area, the cingulate sulcus visual (CSv) area and the posterior insular cortex (PIC). Areas hMT+, CSv, and PIC were equivalently responsive during lateral and antero-posterior GVS while areas VIP and V6 were highly activated during antero-posterior GVS but remained silent during lateral GVS. Using psychophysiological interaction (PPI) analyses, we confirmed that a cortical network including areas V6 and VIP is engaged during antero-posterior GVS. Our results suggest that V6 and VIP play a specific role in processing multisensory signals specific to locomotion during navigation.

Introduction

When navigating through an environment, the human brain performs a complex "sensor fusion" act, integrating visual information (optic flow) with vestibular signals from the inner ear. This process, known as egomotion processing, is fundamental to locomotion. While the visual components of this system are well-mapped, the vestibular contributions—specifically those related to forward and backward movement (antero-posterior motion)—have been less clear.

Most previous research utilized "binaural bipolar" stimulation, which primarily simulates lateral tilt. However, human locomotion is predominantly an antero-posterior activity. This study aimed to identify whether specific cortical networks are dedicated to the vestibular signals generated during actual forward/backward locomotion, distinguishing them from those processing lateral motion.

Methods

The researchers employed a combination of functional magnetic resonance imaging (fMRI) and a specialized form of Galvanic Vestibular Stimulation (GVS) on 13 healthy subjects.

GVS Configurations

To differentiate between motion types, the study utilized three distinct electrode configurations:

  1. Binaural Monopolar (Antero-posterior): Electrodes placed on both the forehead and mastoid processes to simulate forward/backward self-motion.
  2. Binaural Bipolar (Lateral): The standard setup used to create a sensation of lateral tilt.
  3. Sham: A control condition to account for non-vestibular sensations.
Figure 2: Experimental GVS configurations (A-C) and behavioral discrimination results (D).

Regions of Interest (ROIs)

Before the main experimental runs, functional localizers were used to map five motion-sensitive areas: the human middle temporal complex (hMT+), V6, the ventral intraparietal area (VIP), the cingulate sulcus visual area (CSv), and the posterior insular cortex (PIC). Connectivity between these regions was assessed using Psychophysiological Interaction (PPI) analysis.

Key Results

Functional Specialization of V6 and VIP

The most significant finding was the selective activation of areas V6 and VIP. These regions were highly responsive during antero-posterior (AP) stimulation but remained functionally "silent" during lateral stimulation. Specifically, V6 showed significantly higher beta values for AP stimulation compared to lateral ($t(9)=6.99, p < 0.0001$). During lateral GVS, activity in V6 was statistically indistinguishable from the sham condition, suggesting these areas are specialized "locomotion zones."

Multisensory Hubs: CSv and PIC

In contrast to the selectivity of V6 and VIP, areas CSv and PIC acted as universal hubs. CSv showed the strongest overall response to vestibular stimulation regardless of direction. PIC demonstrated high connectivity across conditions, suggesting it plays a central role in integrating diverse multisensory motion signals.

Network Reconfiguration

PPI analysis revealed that antero-posterior stimulation reconfigures the brain's functional architecture. During simulated forward motion, a robust network including V6, VIP, CSv, and hMT+ becomes engaged. Conversely, lateral stimulation primarily strengthens connections between PIC, CSv, and hMT+, effectively leaving the V6/VIP locomotion circuit "out of the loop."

Behavioral Validation

To ensure the GVS effectively induced the intended sensations, subjects performed a discrimination task. As shown in Figure 2D, more than half of the participants (7 out of 13) could significantly discriminate between the two motion directions even at a low current intensity (1mA), confirming the validity of the stimulation protocol.

Discussion and Limitations

The study identifies V6 and VIP as critical components of a dedicated network for processing locomotion-specific vestibular signals. However, the authors note some limitations. The V6 ROI was defined without wide-field retinotopic mapping, meaning it might include parts of the neighboring area V6A. Additionally, the low stimulus intensity (chosen to minimize tactile artifacts) may have resulted in weaker conscious motion perceptions for some subjects, despite clear cortical responses.

Future research will likely focus on combining these specific vestibular stimulations with visual optic flow to explore how the brain resolves sensory conflicts—such as when visual cues indicate forward motion while vestibular cues suggest a lateral tilt.

Summary updated

Abstract

Abstract Visuo-vestibular integration is crucial for locomotion, yet cortical mechanisms involved remain poorly understood. We combined binaural monopolar galvanic vestibular stimulation (GVS) and functional magnetic resonance imaging (fMRI) to characterize the cortical networks activated during antero-posterior and lateral stimulations in humans. We focused on functional areas that selectively respond to egomotion-consistent optic flow patterns: the human middle temporal complex (hMT+), V6, the ventral intraparietal (VIP) area, the cingulate sulcus visual (CSv) area and the posterior insular cortex (PIC). Areas hMT+, CSv, and PIC were equivalently responsive during lateral and antero-posterior GVS while areas VIP and V6 were highly activated during antero-posterior GVS but remained silent during lateral GVS. Using psychophysiological interaction (PPI) analyses, we confirmed that a cortical network including areas V6 and VIP is engaged during antero-posterior GVS. Our results suggest that V6 and VIP play a specific role in processing multisensory signals specific to locomotion during navigation.
BibTeX
@inproceedings{aedojury2020antero,
  author = {Felipe Aedo-Jury and Benoit R. Cottereau and Simona Celebrini and Alexandra Séverac Cauquil},
  title = {{Antero-posterior versus lateral vestibular input processing in human visual cortex}},
  booktitle = {Frontiers in Integrative Neuroscience},
  year = {2020},
  doi = {10.3389/fnint.2020.00043},
}
Figure