| name | eccentricity-confound-eeg-visual-attention-decoding |
| description | EEG-based visual attention decoding methodology addressing the eccentricity confound. Uses spatial attention tasks with controlled stimulus eccentricity and phase-scrambling controls to demonstrate that decodable EEG signals reflect genuine attention modulation, not visual processing confounds. Use when building EEG-based BCI attention decoders, visual attention research, or addressing confounds in neural decoding. |
| version | 1.0.0 |
| metadata | {"hermes":{"source_paper":"Decoding Visual Attention from EEG while Controlling for Visual Confounds (arXiv:2604.14890)","tags":["neuroscience","eeg","visual-attention","decoding","bci","confound-control"]}} |
EEG Visual Attention Decoding with Eccentricity Control
Overview
Addresses the fundamental confound in EEG-based visual attention decoding: whether signals reflect genuine attentional modulation or low-level visual processing differences. Demonstrates that spatial attention can be decoded from EEG above chance, and that the decoded signal is primarily driven by attentional modulation rather than stimulus eccentricity.
Core Contribution
Shows that when controlling for stimulus eccentricity and using phase-scrambled stimuli, the decoded EEG signal reflects genuine spatial attention modulation. This validates EEG-based attention decoding as a reliable BCI paradigm.
Key Findings
| Metric | Value |
|---|
| Attention decoding accuracy | 76% |
| Eccentricity decoding | 55% |
| Phase-scrambled control | Attention signal preserved |
| Conclusion | Attention > eccentricity as driver |
Methodology
def decode_attention(eeg_data, condition):
"""Decode spatial attention from EEG signals."""
pass
Applications
- BCI attention control: Reliable spatial attention-based interfaces
- Visual neuroscience: Disentangling attention from visual processing
- Neurofeedback: Attention-based training systems
- Experimental design: Proper confound control in neural decoding
References
- Original paper: arXiv:2604.14890v1
- Authors: Not specified in abstract
- Published: 2026-04-16