Multimodal Integration of Wearable Sensor Signals for Ambulatory Gait and Posture Analysis

用于动态步态和姿势分析的可穿戴传感器信号的多模态集成

• 批准号: RGPIN-2015-05317
• 负责人: Tung, James
• 金额: $ 1.6万
• 依托单位: University of Waterloo
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2020
• 资助国家: 加拿大
• 起止时间: 2020-01-01 至 2021-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=708638
• 关键词: Multimodal; Integration; Wearable; Sensor; Signals

Wireless, wearable sensor systems (WSS) capable of capturing continuous, objective metrics of gait and postural behavior offer new possibilities for human motor control research, monitoring and diagnosing rehabilitation outcomes for diagnosis, and control systems design of assistive technologies. Major technical challenges that limit the current applicability and efficiency of WSS include: 1) suboptimal accuracy and reliability in capturing detailed gait parameters, 2) few techniques to extract contextual information about the physical environment to interpret data, and 3) few validated approaches to analyze large, longitudinal datasets of gait and posture behaviour. The long-term goal of this research program is to advance the accuracy, reliability, and applicability of WSS for ambulatory gait and posture analysis. The proposed research adopts a multimodal sensor integration approach to advance WSS tools to measure gait and postural control in an ambulatory fashion. The integration of candidate techniques, including spatial mapping and electrophysiological sensors, with widely-used body-worn motion sensors will be explored. To address anatomical and functional heterogeneity, electrophysiological signals will be acquired to provide insight into underlying neuromotor control strategies. The potential for localization and spatial mapping techniques to address the lack of concurrent contextual data will be evaluated. For example, the utility of sensor packages incorporating localization and 3D mapping techniques (e.g., Project Tango by Google, Inc.) to capture environmental information and infer mobility context will be examined. To integrate signals from multiple modalities, sensor fusion techniques drawn from control systems, machine learning, and health informatics fields will be explored. The research will contribute to establishing a centre of WSS expertise to foster internationally recognized research, technical innovation, and human capital for Canada. The proposed advances in WSS capabilities will have a major impact in providing foundational technology for future applications of WSS, principally for uses related to human motor control. Specifically, this program will provide new tools to examine gait and postural control in their natural context and expand the utilization of WSS for applied research (e.g., control of advanced assistive technologies such as exoskeletons, functional electrical stimulators). To meet the projected demands of an exploding mobile health technology (mHealth) market, this program will generate infrastructure for rapid innovation and support the training of highly sought biomedical engineering and health informatics human resources.

无线可穿戴传感器系统(WSS)能够捕获连续、客观的步态和姿势行为指标，为人体运动控制研究、诊断康复结果的监测和诊断以及辅助技术的控制系统设计提供了新的可能性。限制WSS当前适用性和效率的主要技术挑战包括：1)捕获详细步态参数的准确性和可靠性不佳；2)提取有关物理环境的上下文信息以解释数据的技术很少；以及3)分析步态和姿势行为的大型纵向数据集的有效方法很少。这项研究计划的长期目标是提高WSS在步行和姿势分析中的准确性、可靠性和适用性。 建议的研究采用多模式传感器集成的方法，以改进WSS工具，以动态方式测量步态和姿势控制。将探索包括空间测绘和电生理传感器在内的候选技术与广泛使用的人体佩戴的运动传感器的集成。为了解决解剖和功能异质性，将获得电生理信号，以提供对潜在的神经运动控制策略的洞察。将评估本地化和空间测绘技术解决缺乏并发背景数据的潜力。例如，结合了本地化和3D地图技术的传感器包的效用(例如，谷歌公司的Project Tango)为了捕获环境信息并推断移动环境，将对其进行检查。为了集成来自多个模式的信号，将探索来自控制系统、机器学习和健康信息学领域的传感器融合技术。 这项研究将有助于建立一个WSS专门知识中心，以促进加拿大的国际公认的研究、技术创新和人力资本。WSS能力方面的拟议进展将对为WSS的未来应用提供基础技术产生重大影响，这些应用主要用于与人类运动控制有关的用途。具体地说，该计划将提供新的工具，在自然环境中检查步态和姿势控制，并扩大WSS在应用研究中的使用(例如，控制先进的辅助技术，如外骨骼、功能性电刺激器)。为了满足爆炸式增长的移动健康技术(MHealth)市场的预期需求，该计划将为快速创新提供基础设施，并支持对备受欢迎的生物医学工程和健康信息学人力资源的培训。