Neural correlation study of swinging and standing during dual-task walking-A pilot study

• Published in: Chinese Automation Congress (Online) pp. 929 - 934
• Main Authors: Zheng, Lei, Chang, Wenwen, Lu, Jialei, Kong, Weixuan, Yan, Guanghui, Guo, Bin
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.11.2024
• Subjects: Arithmetic; Brain; Brain activity; Character recognition; Dual-task walking; Electrodes; Electroencephalography; Gait phase; Interference; Legged locomotion; Motors; Robots; Time-frequency analysis
• ISSN: 2688-0938
• DOI: 10.1109/CAC63892.2024.10865097

The walking process is often accompanied by other cognitive tasks, which will cause certain interference to the gait. It is important to explore the cognitive characteristics of brain at different stages during dual-task walking for the neural decoding of real gait features. By collecting and analyzing EEG signals of healthy subjects during walking of different cognitive tasks, this study completed the analysis of EEG characteristics during the standing and swinging stages of the whole gait cycle. The time-frequency characteristics of EEG signals on the main electrodes in the locomotor brain, the topological distribution of power spectral density in the swinging and standing phases, and the corresponding brain functional network structure were analyzed. The results showed that the temporal-frequency characteristics and brain functional network connections differed significantly between the non-cognitive and cognitive tasks, and the power spectral characteristics and network topological connections of the EEG signals also showed significant differences between the swinging and the standing phases. However, the power spectral values and functional brain connectivity relationships did not differ between arithmetic operation task and left and right brain task. There was no significant variability in the topological distribution of power spectrum values across brain regions. These results suggest that it is feasible to achieve effective recognition of swing and stand phases within the gait cycle based on EEG features analysis, and that the identification of EEG signals in different walking phases can contribute to the development of brain-computer interface-based rehabilitation and robotic assistive systems.