Long-Range α-Synchronization as Control Signal for BCI: A Feasibility Study

Abstract

Shifts in spatial attention are associated with variations in alpha-band (α, 8–14 Hz) activity, specifically in inter-hemispheric imbalance. The underlying mechanism is attributed to local α-synchronisation, which regulates local inhibition of neural excitability, and fronto-parietal synchronisation reflecting long-range communication. The direction-specific nature of this neural correlate brings forward its potential as a control signal in brain-computer interfaces (BCI). In the present study, we explored whether long-range α-synchronisation presents lateralised patterns dependent on voluntary attention orienting and whether these neural patterns can be picked up at a single-trial level to provide a control signal for active BCI. We collected electroencephalography (EEG) data from a cohort of healthy adults (n = 10) while performing a covert visuospatial attention (CVSA) task. The data shows a lateralised pattern of α-band phase coupling between frontal and parieto-occipital regions after target presentation, replicating previous findings. This pattern, however, was not evident during the cue-to-target orienting interval, the ideal time window for BCI. Furthermore, decoding the direction of attention trial-by-trial from cue-locked synchronisation with support vector machines (SVM) was at chance-level. The present findings suggest EEG may not be capable of detecting long-range α-synchronisation in attentional orienting on a single-trial basis and, thus, highlight the limitations of this metric as a reliable signal for BCI control.SIGNIFICANCE STATEMENTCognitive neuroscience advances should ideally have a real-world impact, with an obvious avenue for transference being BCI applications. The hope is to faithfully translate user-generated brain endogenous states into control signals to actuate devices. A paramount challenge for transfer is to move from group-level, multi-trial average approaches to single-trial level. Here, we evaluated the feasibility of single-trial estimation of phase synchrony across distant brain regions. Although many studies link attention to long-range synchrony modulation, this metric has never been used to control BCI. We present a first attempt of a synchrony-based BCI that, albeit unsuccessful, should help break new ground to map endogenous attention shifts to real-time control of brain-computer actuated systems.

Epistemic Challenges in Neurophenomenology: Exploring the Reliability of Knowledge and Its Ontological Implications

“…Brain-computer interfaces allow for direct communication between the brain and external devices, which can be used to investigate the neural correlates of subjective experiences. For example, BCIs can decode and interpret neural activity patterns associated with attention, perception, and emotional responses [122][123][124], providing insights into the neural mechanisms underlying conscious experiences. BCIs can also improve the accuracy of self-reporting as they provide objective measures of subjective experiences [125,126].…”

Decoding Attentional Task Performance Using Electroencephalogram Signals

“…Gamma band activity, combined with frontal midline theta rhythm in the left and right frontal forehead regions, appears to reflect prefrontal function in working memory tasks [29]. Furthermore, lateralized patterns of band phase coupling between the frontal and parietaloccipital regions have been identified during covert visuospatial attention tasks in healthy young adults [30]. While these findings enrich our understanding of the brain mechanisms underlying attention, further exploration into additional EEG frequency considerations and their potential impact on task performance, particularly lateralized patterns, is essential.…”

Brain–Computer Interface Based on PLV-Spatial Filter and LSTM Classification for Intuitive Control of Avatars

“…The PLV is a metric commonly used to measure synchronization in EEG sensors [40,41] for a BCI because synchronization metrics are less affected by noise than amplitude (this is indicated by a non-random phase or phase difference distribution [42]). Noise in an EEG signal can directly affect the signal, while the phase of the two signals can remain relatively unchanged by amplitude noise since this type of noise generally tends to affect the magnitude of the signal more than its relative phase.…”