A Multimode Markerless Gait Motion Analysis System Based on Lightweight Pose Estimation Networks

作者：Weining Li, Shimeng Wang, Jiaxin Lei, Xuecheng Wang, Liu Wang, Kuntao Chen, Tieyuan Zhang, Yanjun Guan, Zhe Zhao, Lan Yin, Milin Zhang · 发表于：2022 IEEE Biomedical Circuits and Systems Conference (BioCAS) · 年份：2022 · DOI：10.1109/biocas54905.2022.9948540 · 被引用次数：3 · 研究领域：Robotic Locomotion and Control、Gait Recognition and Analysis、Muscle activation and electromyography studies

Gait motion analysis for animal models plays an important role in experimental medical science to better study neurological disorders. To perform a quantitative assessment of neural diseases, this paper proposed a multi-mode markerless gait motion analysis with synchronous acquisition of neural signals and dual-view gait video recording. To achieve the goal of real-time online data analysis and on-device implement, the design of lightweight pose estimation networks was presented with the output of nine anatomical features related to gait motion patterns. Fusion analysis method for collected multimode information of gait parameters and EMG/ENG signals was introduced. The proposed gait motion analysis system workflow was verified by experiments upon two groups of rats including normal rats and rats of nerve injury.