Synchronization-based fusion of EEG and eye blink signals for enhanced decoding accuracy

Alyan, Emad; Arnau, Stefan; Reiser, Julian Elias; Wascher, Edmund · 2024 · Crossref

DOI: 10.1038/s41598-024-78542-9

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of decoding brain activity during dynamic locomotor tasks, where traditional single-modality electroencephalography (EEG) often lacks sufficient precision. The authors argue that integrating EEG with ocular metrics, specifically involuntary eye blinks, offers a promising avenue for understanding cognitive processes, as blinks reflect neural mechanisms and attentional shifts. However, previous attempts at multimodal fusion have been hindered by inconsistent temporal alignment between EEG signals and blink events. To resolve this, the paper introduces a novel methodology called pcEEG+, which fuses EEG principal components with temporally synchronized eye blink data to enhance decoding accuracy. The research utilized data from 35 participants performing auditory tasks while engaging in three locomotor conditions: standing, walking, and navigating obstacles. EEG data were recorded using a 30-electode mobile system, and eye blink signals were extracted directly from the EEG data using independent component analysis rather than external sensors. The core methodological innovation involves calculating the temporal offset between the peak of the blink amplitude and the peak of the global field power (GFP) in the EEG signal. This offset is used to shift the blink signal, ensuring precise synchronization with neural activity. The synchronized blink data is then concatenated with EEG principal components (derived via Principal Component Analysis) or canonical variables (derived via Canonical Correlation Analysis) to create a fused feature set for multivariate pattern analysis. The results demonstrated that the pcEEG+ method significantly improved decoding accuracy for locomotor tasks, reaching 78% in some conditions. This performance surpassed standalone pcEEG methods by 7.6% and standalone synchronized blink methods by 22.7%. Temporal generalization matrices confirmed the consistency of the pcEEG+ approach across different tasks and time points. To validate the robustness of the method, the findings were replicated using two distinct driving simulator datasets involving proactive and reactive driving tasks. The replication confirmed that the synchronization-based fusion effectively generalized beyond the initial locomotor context, maintaining superior decoding performance compared to single-modality analyses. The significance of this work lies in its demonstration that precise temporal synchronization is critical for effective multimodal fusion in neuroscience. By treating eye blinks as informative event markers rather than mere artifacts, the study provides a deeper insight into brain-ocular dynamics during complex movements. The pcEEG+ method offers a reliable framework for decoding cognitive states in real-world scenarios, addressing a key gap in understanding how the brain processes information during multitasking and physical activity. This approach enhances the interpretative depth of EEG data and supports the development of more accurate brain-computer interfaces and cognitive monitoring tools.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-09; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• drowsiness detection algorithms

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: physiological data