.png){kind=icon}
.png){kind=icon}
Return
Adaptive Fuzzy-Convolution and TSK-Guided Attention for Interpretable EEG MI Decoding
.png){kind=icon}
PRE
AI
DOI：10.1109/TFUZZ.2025.3643920.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Brain–computer interface (BCI) technology enables direct communication between the brain and external devices via noninvasive methods and holds significant potential in neuroengineering, rehabilitation, and human–computer interaction. However, decoding motor imagery (MI) from electroencephalogram (EEG) signals remains challenging due to these signals’ nonstationary characteristics and the limited interpretability of existing deep learning models. In this article, we propose a novel hierarchical collaborative fuzzy network for interpretable EEG-based MI decoding. We introduce an adaptive fuzzy temporal convolutional network that employs dynamic fuzzy kernels within causal convolutions to extract robust temporal features from EEG signals. In addition, we design a fuzzy attention-guided Takagi–Sugeno–Kang architecture that achieves a tighter integration between feature extraction and fuzzy inference through a novel fuzzy feedback loop, thereby improving the discriminability of extracted features. Extensive experiments on the BCI Competition IV-2a, IV2b, and OpenBMI datasets, under both subject-dependent and cross-subject evaluation paradigms, demonstrate that the proposed model outperforms state-of-the-art methods in classification accuracy and Cohen’s kappa. Furthermore, we provide multilevel interpretability analyses, from macro to microperspectives, elucidating the model’s decision-making processes and highlighting the advantages of our collaborative reasoning framework over conventional cascaded approaches.
Keywords:
Brain–computer interface (BCI)
electroencephalogram (EEG)
hierarchical collaborative fuzzy network
interpretability
motor imagery (MI)
Takagi–Sugeno–Kang (TSK) fuzzy system
Journal
IF:
11.9
Papers:
4.9K
Citations:
2.9W