Logo image
Back
Towards Adaptive ERP-Based BCIs: EEG and fNIRS Guided Flashing Strategies to Account for Cognitive Fatigue
Conference paper   Open access

Towards Adaptive ERP-Based BCIs: EEG and fNIRS Guided Flashing Strategies to Account for Cognitive Fatigue

Onur Erdem Korkmaz, Tolga Turay, Hasan Kilickaya, Hasan Ayaz and Riccardo Poli
2025 16th International Conference on Electrical and Electronics Engineering (ELECO), pp 1-5
27 Nov 2025
DOI: https://doi.org/10.1109/ELECO69582.2025.11329354
url
https://avesis.atauni.edu.tr/publication/details/7a8b347b-501e-48d8-ab65-20c60413ec4d/oaiView
Submitted Open CC BY V4.0

Abstract

Accuracy Faces Fatigue Functional near-infrared spectroscopy Problem-solving Real-time systems Sensitivity Signal to noise ratio Time factors Electroencephalography
This pilot study presents a hybrid EEG-fNIRS framework to jointly assess speller performance and cognitive fatigue. Speller paradigms often face a trade-off: increasing stimulus flash repetitions improves ERP signal-to-noise ratio (SNR) but prolongs task duration, while fewer flashes reduce efficiency under fatigue. To address this limitation, EEG and fNIRS signals were recorded simultaneously during speller and cognitive task, arithmetic problem-solving blocks. Behavioral findings showed a decline in performance of the cognitive task, with fewer correct answers and longer response times across blocks, suggesting increased fatigue. ERP analyses revealed overall strong target responses but with attenuated amplitudes over time, especially in parietal and occipital channels. Classification confirmed these patterns: target vs. non-target discrimination reached 98.6% accuracy with three EEG channels, cognitive block classification achieved 92.4% using EEG, and 87.4% using fNIRS. These results demonstrate the feasibility of hybrid EEG-fNIRS systems for monitoring user states and provide a foundation for adaptive speller BCIs that dynamically adjust stimulus repetitions.

Metrics

1 Record Views

Details

Logo image