步态对于高龄人群而言是重要的身体指标，因为他们容易受到虚弱或跌倒的风险。智能可穿戴设备不受时间地点的限制，为步态评估提供了低成本的创新方案。 本文采用智能设备开展高龄人群的步态评估研究，研究框架由步态评估模型、步态系统、步态数据库组成。本文采用智能设备(智能手环、手机)搭建步态系统，在老人的自然状态下采集步态数据；分别在四肢和后腰接近人体重心处佩戴智能设备，邀请社区老人参加步态测试，建立单任务、认知多任务和快速走下的步态模型；采用以用户为中心的设计流程，了解老人和医学专业人士对于工具的意见和建议，设计开发了可用于日常步态测试的手机应用“口袋步态测试”，并验证应用在北京和重庆老人中的可用性。 研究结果表明：基于右脚踝的智能手环传感器，加速度均方根随着速度增加而显著增加；与年轻人相比，老人在前后和左右方向上呈现较低幅值的加速度模式，而年轻人和老人的步伐规则性和变异性没有显著差异。基于后腰部的智能手环步态系统，虚弱前期老人的速度，加速度均方根，垂直方向加速度幅值变异性和步伐规则性显著小于非虚弱老人；表明智能手环可定量检测通过视觉或功能测试不易观察到的虚弱相关的步态特征(例如，加速度均方根，步伐规则性)，增强医护人员的步态评估效力。本研究设计的手机应用“口袋步态测试”，通过手机内置的加速度传感器记录步态数据和生成评估报告，包含步态测试、查看图表、查看报告、发送四个功能模块，定位为居家步态评估。研究使用该应用在实验室环境中采集了140位老人的步态数据，建立了中国老人的步态数据库。用户评估表明，老人和医学专业人士对步态评估工具均持积极态度。北京的老人对手机应用的可用性评价显著优于重庆老人，而60岁组老人对应用的可用性评价显著优于80岁组老人。 至此，本文从步态评估模型的建立、基于可穿戴设备的步态系统的开发、步态数据库的建立三个方向，完成了智能设备步态评估框架的建立，为未来使用可穿戴设备开展居家步态测试提供了理论参考，对于进一步实践科技化养老有重要的指导意义。 关键词：步态；科技养老；以用户为中心的设计；医疗信息学；移动医疗

 Gait assessment is useful for older adults because they are vulnerable to frailty or fall risk. Unrestricted by time and place, wearable devices provide cost-effective solutions for gait assessment among older adults. In this thesis, wearable devices with built-in accelerometers were deployed for gait assessement among older adults. The research framework consisted of gait assessment model, system, and database. This thesis used cost-effective smart devices (smart bracelet and smartphone) to build gait assessment systems, collecting gait data in the natural statues of the older adults. This thesis established reference gait parameters of older Chinese adults using smart bracelets under single task, cognitive dual task and fast walking. This thesis adopted a user-centered design process to understand the requirements of the older adults and medical personnels on the gait assessment tools. This thesis then designed and developed a mobile phone based gait assessment application called “Pocket Gait”, and verified the usability of the application among older adults from Beijing and Chongqing. Results indicated that the root mean square (RMS) acceleration increased significantly as the walking pace increased based on the right ankle sensor. Older participants usually presented a lower magnitude of acceleration patterns in the anteroposterior (AP) and mediolateral (ML) direction compared with the younger participants, while the stride regularity and variability were not significantly different. Based on the gait parameters derived from the smart bracelet sensor in the L3 region of the back, pre-frail older adults showed significantly decreased speed, acceleration RMS, vertical (VT) amplitude variability and VT step regularity compared with their non-frail counterparts; this will improve clinicians/community caregivers to identify older adults with frailty-related decline of gait parameters which are not easily observed by vision or functional tests. The mobile phone application “Pocket Gait” was positioned as a homecare gait assessment tool. It could record gait data and generate gait assessment report based on its built-in accelerometer. It includes four modules: gait test, viewing graph, viewing report and sending data. User evaluation indicated that both the older adults and medical personnels held positive attitude towards the gait assessment tools. Older adults from Beijing scored significantly higher System Usability Scale (SUS) compared with older adults from Chongqing. The older adults between 60 and 69 years had siginicantly higher SUS score than those over 80 years of age. In conclusion, this thesis developed a framework consisting of gait assessment model, system (“Pocket Gait”), and empirical gait database based on wearable devices. It served as theoretical reference for further practice of technological elderly care. Key words: gait; technological elderly care; user-centered design; medical informatics; m-health products