Engaging the Patient into the Virtual Care Loop: An E-Health Framework using Data and Knowledge-Driven Methods for Ambient Sensor-Based Health Activity Monitoring to Remotely Determine Patient's Health and Functional Status

点击翻译按钮获取中文摘要

In Canada, there is a recognition of the need to have a healthcare system which provides at--home care, especially to the elderly, chronic and mental health patients. The recent covid-19 pandemic further emphasized the need and utility of virtual care programs that offer remote therapeutic consultations to patients in the safety of their homes. Our intent is to provide unobtrusive and privacy-preserving remote monitoring of seniors, chronic and mental health patients' living at home, to assist them self-manage their health, maintain their autonomy, lifestyle and comfort whilst living a healthy life at their homes. This research proposes a novel e--Health based virtual care platform that employs ambient intelligence to provide assistive healthcare by remotely monitoring a patient's health in a home setting. Ambient intelligence refers to smart environments that use environmental sensors to collect location and movement data of individuals to help them perform daily-life activities by providing guidance and reminders. We are proposing a novel virtual care approach based on the concept of Activity Recognition (AR)-i.e. monitoring the targeted individual's daily life activities to determine health and function. It has been noted that when health is affected, the ability to carry out routine daily-life activities is also affected, so by identifying variations in daily-life activities over time we will be able to detect decline in health and deliver timely assistive healthcare interventions. This research will pursue scientific investigations in remote sensing, knowledge-- and data--driven decision support and e--Health systems to develop innovative unobtrusive AR methods for remote monitoring of a patient's daily activities. The AR methods will be used to generate digital biomarkers to assess the ability of an individual to perform daily-life activities. For remote sensing, we will use contactless sensors to collect the patients' activity data. We will develop a novel hybrid AR solution, where knowledge--driven methods will be used to represent the patient's home environment and daily-life activity models, and data--driven methods to recognize the patients' daily-life activities. Based on the recognized activities, we will generate digital biomarkers to determine the patient's health status. The AR and health assessment methods will be implemented as an e-Health virtual care platform for healthcare providers to remotely monitor their patient's health status. This project's significance is an e-Health platform for virtual care, enabling seniors and patients with dementia, mental health and chronic conditions to safely live in their homes. This research is timely and relevant given the increase in elderly Canadians and in light of the loss of life due to COVID-19 in long-term care facilities. The project will develop novel AR methods to identify the activities of a specific person in a multi-person home and personalize activity models over time.