Yoo Sejin Yi, Gille Emma A, Gomez Jorge, Rocon Eduardo, Martinez-Caballero Ignacio, Van Asseldonk Edwin, Bayon Cristina
IEEE Int Conf Rehabil Robot. 2025 May;2025:687-692. doi: 10.1109/ICORR66766.2025.11062985.
Ankle Foot Orthoses (AFOs) are devices designed to assist or rehabilitate gait. The stiffness of the AFO is an important parameter that needs to be tailored to the enduser's characteristics. However, the lack of clinical guidelines for optimizing AFO stiffness has led to inconsistent outcomes and low user satisfaction in existing experimental approaches. This inconsistency may stem from two reasons: (1) the limited number of variables considered by these methods, which may fail to capture the full complexity of gait; and (2) the variability in the relationship between stiffness and gait variables among users. To address these challenges, we propose a user-tailored optimization framework that identifies and incorporates relevant gait variables based on individual needs and preferences. In this pilot study, we explore key performance variables in two participants with cerebral palsy (CP), analyzing the results using XGBoost and SHAP values. Our findings highlight the importance of multiple performance variables in capturing gait complexity and reveal that the most relevant variables differ between participants. This study underscores the potential of personalized approaches to improve AFO optimization and user outcomes.
踝足矫形器(AFO)是用于辅助或康复步态的装置。AFO的刚度是一个重要参数,需要根据最终用户的特征进行调整。然而,缺乏优化AFO刚度的临床指南导致现有实验方法的结果不一致且用户满意度较低。这种不一致可能源于两个原因:(1)这些方法考虑的变量数量有限,可能无法捕捉步态的全部复杂性;(2)用户之间刚度与步态变量关系的变异性。为应对这些挑战,我们提出了一个用户定制的优化框架,该框架根据个人需求和偏好识别并纳入相关的步态变量。在这项初步研究中,我们在两名脑瘫(CP)患者中探索关键性能变量,使用XGBoost和SHAP值分析结果。我们的研究结果突出了多个性能变量在捕捉步态复杂性方面的重要性,并揭示了不同参与者之间最相关的变量有所不同。这项研究强调了个性化方法在改善AFO优化和用户结果方面的潜力。