Intelligent Processing Systems for Zero-Effort Health Smart Homes

零努力健康智能家居的智能处理系统

• 批准号: RGPIN-2020-05882
• 负责人: Armanfard, Narges
• 金额: $ 2.04万
• 依托单位: McGill University
• 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=718057
• 关键词: Intelligent; Processing; Systems; Zero; Effort

Physiological parameters as well as daily activity patterns are essential metrics when monitoring health and well-being of health smart home (HSH) occupants such as elderly people. The aim of HSH is not only to reduce pressure on the health system but also to provide timely e-health services to individuals wishing to live in their own homes and maintain their independence. Zero-effort HSH, which is a type of HSH, aims to realize the HSH goals such that neither active interaction or compliance from the user is required, and the user doesn't need to wear any sensor or device. In fact, in Z-HSH, technology is embedded into the common household items and doesn't interfere with the daily activities of the smart home occupants. This research program will develop Z-HSH's intelligent systems required for unobtrusive physiological parameters estimation and daily activity recognition, incorporating advanced machine learning techniques. Essential physiological parameters, including beat-to-beat systolic and diastolic blood pressures, heart rate, heart beat location and blood oxygen saturation, will be estimated through an intelligent combination of raw physiological data recorded by a zero-effort smart home-based package (Z-SHP), during the common daily activity sitting and watching TV. The Z-SHP consists of a smart chair, a smart tile and two fixed location RGB cameras that unobtrusively measure electrocardiogram, photoplethysmogram, ballistocardiogram and face and hand videos of the user while he/she is sitting on the smart chair, in front of the cameras (watching TV), with his/her feet placed on the smart tile. The intelligent system for daily activity recognition will be based on an activity segmentation framework as well as a multi-task learning framework that stitches human pose, attention control and daily activity recognition in the benefit of the latter one. This research program will impact on self-management, quality of life, and health outcomes of smart home occupants who would like to go about their usual daily routines within their own homes, with the required measurements being automatically collected (and delivered to the appropriate stakeholders such as caregivers and health centers) as they interact with common everyday objects. So it will impact the key clinical outcomes, such as a reduction in (re)-admission rates, hence offering improvements to the Canadian economy as a whole.

生理参数以及日常活动模式是监测健康智能家居（HSH）居住者（如老年人）的健康和福祉时的重要指标。HSH的目标不仅是减轻对卫生系统的压力，而且还为希望住在自己家里并保持独立的个人提供及时的电子卫生服务。零努力HSH是HSH的一种类型，旨在实现HSH目标，使得既不需要来自用户的主动交互或顺从，也不需要用户佩戴任何传感器或设备。事实上，在Z-HSH中，技术被嵌入到普通的家居用品中，不会干扰智能家居居住者的日常活动。 该研究计划将开发Z-HSH的智能系统，用于非侵入性的生理参数估计和日常活动识别，并结合先进的机器学习技术。基本的生理参数，包括心跳到心跳的收缩压和舒张压，心率，心跳位置和血氧饱和度，将通过零努力智能家居包（Z-SHP）记录的原始生理数据的智能组合进行估计，在常见的日常活动坐着和看电视。Z-SHP由一个智能椅子、一个智能瓷砖和两个固定位置的RGB摄像头组成，当他/她坐在智能椅子上、在摄像头前面（看电视）、他/她的脚放在智能瓷砖上时，这些摄像头可以不显眼地测量心电图、光电容积描记图、心冲击图以及用户的面部和手部视频。用于日常活动识别的智能系统将基于活动分割框架以及多任务学习框架，该框架将人类姿势，注意力控制和日常活动识别结合起来，以利于后者。 该研究计划将影响智能家居居住者的自我管理，生活质量和健康结果，这些居住者希望在自己的家中进行日常生活，当他们与常见的日常物品互动时，自动收集所需的测量结果（并交付给适当的利益相关者，如护理人员和健康中心）。因此，它将影响关键的临床结果，例如降低（再）入院率，从而改善加拿大整体经济。