TopoCog: EEG Sensor-Based Cognitive State Recognition in Interactive Environments Using Topological Data Analysis
IEEE Sensors Letters · 2026-12 · 10 (12) · 6013104-6013104
DOI: 10.1109/lsens.2026.3727329 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Sensors Letters, 2026-12. Verfasst von Kumar Gaurav und 2 weiteren Autor:innen. Die Arbeit nennt 16 Literaturangaben.
Abstract (Originalsprache)
The recognition of users' cognitive states (CS) during video gameplay is a key challenge in advancing human-computer interaction and affective computing. This letter presents a novel approach for classifying cognitive states such as engagement, interest, stress, and excitement using electroencephalography (EEG) sensor signals acquired during gameplay. The proposed method is based on topological data analysis (TDA), which enables the extraction of robust and discriminative features by capturing the intrinsic geometric and topological structures of multi-channel EEG data. Specifically, persistence diagrams are computed from EEG time-series to encode multi-scale neural activity patterns, reducing reliance on conventional handcrafted features.The effectiveness of the proposed framework is evaluated on the publicly available MultiPENG dataset, which includes annotated EEG recordings corresponding to different cognitive states elicited during gaming sessions. Experimental results demonstrate that the proposed approach achieves an average accuracy of 74.51±0.32% for multi-class CS recognition. These results outperform several existing state-of-the-art methods, highlighting the advantage of leveraging topological representations for modeling complex brain dynamics. Furthermore, the experimental analysis indicates that the proposed method provides consistent and reliable performance across varying cognitive conditions. The findings suggest that TDA-based feature extraction is a promising direction for EEGbased cognitive state recognition. The proposed approach has significant potential for applications in adaptive gaming environments, user experience evaluation, and real-time human-computer interaction (HCI) sensor-based systems.
Zitieren
APA
Gaurav, K., Sharma, N., & Bollu, T. K. R. (2026). TopoCog: EEG Sensor-Based Cognitive State Recognition in Interactive Environments Using Topological Data Analysis. IEEE Sensors Letters, 10(12), 6013104-6013104. https://doi.org/10.1109/lsens.2026.3727329
Vancouver
Gaurav K, Sharma N, Bollu TKR. TopoCog: EEG Sensor-Based Cognitive State Recognition in Interactive Environments Using Topological Data Analysis. IEEE Sensors Letters. 2026;10(12):6013104-6013104. doi:10.1109/lsens.2026.3727329
BibTeX
@article{Gaurav2026,
author = {Gaurav, Kumar and Sharma, Neha and Bollu, Tharun Kumar Reddy},
title = {TopoCog: EEG Sensor-Based Cognitive State Recognition in Interactive Environments Using Topological Data Analysis},
journal = {IEEE Sensors Letters},
year = {2026},
volume = {10},
number = {12},
pages = {6013104--6013104},
doi = {10.1109/lsens.2026.3727329}
}
EUVIMED