Working-memory load decoding model inspired by brain cognition based on cross-frequency coupling

• Published in: Brain research bulletin Vol. 221; p. 111206
• Main Authors: Zhang, Jing, Tan, Tingyi, Jiang, Yuhao, Tan, Congming, Hu, Liangliang, Xiong, Daowen, Ding, Yikang, Huang, Guowei, Qin, Junjie, Tian, Yin
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.02.2025; Elsevier
• Subjects: Brain - physiology; Cognition - physiology; Cross-frequency coupling; Delayed matching-to-sample; EEG decoding; Electroencephalography - methods; Humans; Interpretability; Memory, Short-Term - physiology; Models, Neurological; Neural Networks, Computer; Sinc convolution layer; Working memory; Cross-frequency coupling; Delayed matching-to-sample; EEG decoding; Sinc convolution layer; Working memory; Interpretability
• ISSN: 0361-9230; 1873-2747; 1873-2747
• DOI: 10.1016/j.brainresbull.2025.111206

Working memory, a fundamental cognitive function of the brain, necessitates the evaluation of cognitive load intensity due to limited cognitive resources. Optimizing cognitive load can enhance task performance efficiency by preventing resource waste and overload. Therefore, identifying working memory load is an essential area of research. Deep learning models have demonstrated remarkable potential in identifying the intensity of working memory load. However, existing neural networks based on electroencephalogram (EEG) decoding primarily focus on temporal and spatial characteristics while neglecting frequency characteristics. Drawing inspiration from the role of cross-frequency coupling in the hippocampal region, which plays a crucial role in advanced cognitive processes such as working memory, this study proposes a Multi-Band Multi-Scale Hybrid Sinc Convolutional Neural Network (MBSincNex). This model integrates multi-frequency and multi-scale Sinc convolution to facilitate time-frequency conversion and extract time-frequency information from multiple rhythms and regions of the EEG data with the aim of effectively model the cross-frequency coupling across different cognitive domains. Due to its unique structural design, the proposed model proficiently extracts features in temporal, frequency, and spatial domains while its feature extraction capability is validated through post-hoc interpretability techniques. On our collected three-class working memory dataset, the proposed model achieved higher classification accuracy compared to other state-of-the-art methods. Furthermore, by analyzing the model’s classification performance during different stages of working memory processes, this study emphasizes the significance of the encoding phase and confirms that behavioral response does not accurately reflect cognitive load. •MBSincNex is a lightweight and interpretable model that efficiently extracts EEG temporal-spatial-frequency features, providing cognitive physiological insights and feature visualization.•Inspired by cognitive mechanisms, the model employs multi-band and multi-scale Sinc convolution to detect cross-frequency coupling features, which play a functional role in WM across multiple brain areas.•By incorporating dilation convolution, MBSincNex broadens the receptive field, enhancing the perception of global information. During training, it combines center loss and cross-entropy loss to achieve compact intra-class features and highly discriminative inter-class features.•TA performance comparison across different stages of WM tasks highlights the critical role of early stages in WM functionality. Furthermore, classification based on WM load labels derived from behavioral responses reveals that behavioral responses do not accurately reflect cognitive load levels.