Development of novel wearable-based IoT system for monitoring changes in mobility, cognition and daily activities in community-dwelling older adults

与Bigmotion Technologies的合作项目将完善，验证和确定其可穿戴移动传感器系统与生活在家庭环境中的体弱老年人的可用性，并开发一种新的预测建模方法，以整合来自多个传感器的“大数据”，以早期识别健康状况的变化，并预测不良健康事件。这种新颖的方法将联合收割机结合多个测量领域的多个输入的预测能力，例如睡眠模式、日常活动模式以及步态和移动参数。拟议的可穿戴监测系统包括Android智能手表，其中包含几种仪器（即加速度计、陀螺仪、磁力计和气压计以及蓝牙低功耗芯片组），可与位置传感器结合使用。该项目将首先完善和验证这一新型系统的能力，通过结合用户反馈并将其与当前的行业标准进行比较，来测量日常活动模式，睡眠模式，步态速度和家庭环境中的跌倒检测。该系统已在实验室中进行了验证，以确保其能够提供准确可靠的移动数据以及位置感测数据和睡眠质量，但尚未在家庭中进行验证。第二，本项目将创新性地将连续和间歇收集的家庭综合数据联合收割机结合起来。通过确定多变量连续家庭传感器数据之间的关系，（例如活动性和日常活动）和认知、活动性和患者报告的不良健康事件的结果，仅基于连续数据的现代计算方法将提供对增加的健康风险的早期识别，例如谵妄、徘徊、福尔斯，以及认知能力的变化或下降，从而能够及时实施适当的预防策略，以减少威胁独立性和生活质量的不良健康事件的发生率。

This collaborative project with Bigmotion Technologies, will refine, validate and determine the usabilityof its wearable mobility sensor system with frail older adults living in the home setting, as well as develop a novel predictive modelling approach to integrate "big data" from its multiple sensors to provide early identification of changes in health status as well as a predict adverse health events. This novel approach will combine the predictive power of multiple inputs across many domains of measurement such as sleep patterns, daily activity patterns, as well as gait and mobility parameters. The proposed wearable monitoring system includes Android smartwatch housing several instruments (i.e. accelerometer, gyroscope, magnetometer, and barometer, and Bluetooth Low Energy chipset) to be used in conjunction with location sensors. The project will first refine and validate the capacityof this novel system to measure daily activity patterns, sleep patterns, gait velocity and fall detection in the home setting by incorporating user feedback and comparing it against current industry standards. The system has been validated in laboratory, to ensure its ability to provide accurate and reliable mobility data and as well as position sensing data and sleep quality, but has not been validated in the home. Second, this project will innovatively combine in-home integrated data collected both continuously and intermittently. By determining the relationship between multivariate continuous in-home sensor data (e.g. mobility and daily activities) and outcomes of cognitive, mobilityand patient-reported adverse health events measured intermittently with conventional assessments, modern computational methods based solely on continuous data will provide early identification of increased health risks, such as delirium, wandering, falls, and variations or declines in cognition allowing for timely implementation of appropriate prevention strategies to mitigate the incidence of adverse health events that threaten independence and quality of life.