文献类型：期刊文献

英文题名：Multiscale Pooling Spatial-Temporal Attention Network: Elevating Cross Session and Small Sample Decoding in Motor Imagery Brain-Computer Interfaces

作者：Chen, Weijie[1];Daly, Ian[2];Chen, Yixin[1];Wu, Xiao[1];He, Xinjie[1];Wang, Xingyu[1];Cichocki, Andrzej[3,4,5];Jin, Jing[1,6]

机构：[1]East China Univ Sci & Technol, Key Lab Smart Mfg Energy Chem Proc, Minist Educ, Shanghai 200237, Peoples R China;[2]Univ Essex, Sch Comp Sci & Elect Engn, Brain Comp Interfacing & Neural Engn Lab, Colchester CO4 3SQ, Essex, England;[3]Nicolaus Copernicus Univ, Dept Informat, PL-87100 Torun, Poland;[4]Polish Acad Sci, Syst Res Inst, PL-01447 Warsaw, Poland;[5]RIKEN, Brain Sci Inst, Lab Adv Brain Signal Proc, Wako 3510198, Japan;[6]East China Univ Sci & Technol, Sch Math, Shanghai 200237, Peoples R China

年份：2026

卷号：56

期号：4

起止页码：2225

外文期刊名：IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS

收录：;EI(收录号：20260319933422);WOS:【SCI-EXPANDED(收录号:WOS:001663456800001)】；

基金：This work was supported in part by the Brain Science and Brain-like Intelligence Technology-National Science and Technology Major Project under Grant 2022ZD0208900; in part by the National Natural Science Foundation of China under Grant 62176090; in part by Shanghai Municipal Science and Technology Major Project under Grant 2021SHZDZX; and in part by the Project of Jiangsu Province Science and Technology Plan Special Fund in 2022 (Key Research and Development Plan Industry Foresight, Fundamental Research Fund for the Central Universities JKH01241605 and Key Core Technologies) under Grant BE2022064-1.

语种：英文

外文关键词：Feature extraction; Electroencephalography; Decoding; Brain modeling; Accuracy; Data mining; Motors; Deep learning; Attention mechanisms; Machine learning algorithms; Brain-computer interfaces (BCIs); cross session; motor imagery (MI); small sample

摘要：Motor imagery (MI) is one of the most widely used paradigms in brain-computer interfaces (BCIs), known for its ability to trigger changes in brain activity without the need for an external "cue" stimulus. This unique characteristic has attracted significant attention from neuroscientists and researchers in fundamental science. However, compared to P300 and steady-state visual evoked potential (SSVEP), neural activity related to MI tends to be less stable and exhibits substantial variability between individuals. Consequently, accurately decoding MI, using both traditional machine learning and deep learning, has proven to be a considerable challenge. Moreover, given the difficulty of acquiring electroencephalography (EEG) data and the high data demands of deep learning, enhancing the accuracy of MI decoding with limited sample sizes remains a pressing issue that urgently needs to be addressed. This article addresses the challenges mentioned above by introducing a novel deep neural network designed for accurate MI decoding, which is designed to be effective with both small-sample sizes and larger datasets. This network, named the multiscale pooling spatial-temporal attention network (MPSTANet), integrates mix pooling techniques with spatial-temporal attention mechanisms. MPSTANet first employs local and global spatial attention, along with multiscale temporal attention, to thoroughly extract spatial-temporal information from EEG signals. Next, MPSTANet utilizes feature fusion and the proposed mix pooling technique to preserve as much of the extracted spatial-temporal information as possible. Finally, channel interaction attention (CIA) and 3-D weight attention (3-DWA) are employed to recalibrate the weights of the fused channels and spatial-temporal features, respectively. To validate the performance of our proposed MPSTANet model, we conducted experiments on four public datasets, including both small-sample sizes and subject-independent scenarios. MPSTANet achieved cross-session decoding accuracies of 84.82%, 72.92%, 88.20%, and 46.54% on the BCI Competition IV 2a dataset, the Open BMI dataset, the BCI Competition IV 2b dataset, and the PhysioNet dataset, respectively. Furthermore, MPSTANet demonstrated a significant lead compared to other deep learning models in both small-sample and subject-independent experiments. These results demonstrate the robustness of MPSTANet in MI decoding and its promising potential for BCI applications.