2023, Vol. 27 ›› Issue (8): 1257-1263

Methods and application of gait analysis of lower limbs after stroke

Yu Wenqiang1, Ren Fuchao1, Shi Guohong2, Xu Yuanjing3, Liu Tongyou2, Xie Youzhuan2, Wang Jinwu2, 3   

Abstract: BACKGROUND: Stroke is one of the major diseases endangering the health of Chinese citizens, characterized by high morbidity, disability, mortality and recurrence. Gait dysfunction or impairment is considered to be one of the most common and devastating physiological consequences of stroke. Active lower extremity rehabilitation can promote the recovery of gait function, improve the ability to perform activities of daily living, and improve the quality of life of patients.
OBJECTIVE: To review and summarize research findings applicable to post-stroke gait quantification and analysis, with a focus on the latest gait analysis system technologies, post-stroke gait data processing and analysis techniques, and their feasibility and potential value in the clinical setting.
METHODS: Search terms included “stroke, gait analysis, assessment, lower limb, spatiotemporal, kinematics, kinetics, plantar pressure, electromyography (EMG), machine learning, statistics” in Chinese and English. CNKI and PubMed databases were searched for relevant articles published from January 2000 to December 2021. A few of classical distant documents were also included. The initial screening was performed by reading the titles and abstracts of the articles. Duplicate studies, low-quality journals and irrelevant documents were excluded and 62 articles were finally included for review.
RESULTS AND CONCLUSION: Traditional qualitative gait analysis is primarily based on observational gait and is therefore subjective and largely influenced by the observer’s experience. Quantitative gait analysis, on the other hand, provides measured parameters with good accuracy and repeatability for diagnostic and comparative assessment throughout the rehabilitation process. The rapid development of smart wearable technologies and artificial intelligence is increasingly drawing more attention to gait research. Although their use in clinical settings is not yet well exploited, these tools are expected to change the status quo in stroke gait quantification by providing the means to acquire, store, and analyze multifactorial complex gait data while capturing its nonlinear dynamic variability and providing the valuable benefit of predictive analysis. Gait analysis is often characterized by a number of specific parameters that describe normal or abnormal gait, including temporal, kinematic, kinetic, and electromyographic parameters. Understanding the factors associated with walking function can help clinicians and researchers identify gait-related parameters that should focus on when assessing walking function. As conventional statistical methods are no longer adequate to handle the large data volumes with high heterogeneity and complexity generated by instrumented gait analysis, and gait analysis involves a large number of interdependent parameters that are difficult to interpret due to the large amount of data and their interrelationships. To simplify the evaluation, the application of machine learning to gait analysis is a promising solution.  
Key words: stroke, gait analysis, gait parameter, statistics, machine learning, rehabilitation, diagnosis, assessment