CDI - Type II: Transforming Community-Based Elder Care through Heterogeneous Activity Sensing Analytics

随着人们因多种慢性病而寿命延长，增加了他们对医疗保健服务的需求，医疗保健系统面临着严峻的挑战。解决人口老龄化的需求需要通过持续监测、早期发现不良事件和模式以及早期干预来改进基于社区的预防性护理方法。该项目通过开发一个用于家庭活动监测和分析的灵活工具的变革管道来应对这一挑战。其主要目标是实现灵活的传感器平台和工具的家庭部署，以便具有不同技能和信息需求的用户灵活分析传感器数据。为了实现这一目标，该项目侧重于三个具体挑战。首先，该项目正在开发新的平台，用于根据不同的老年人需求部署不同的家庭传感器组合。其次，该项目正在创建新的工具，将原始传感器数据转换为老年人活动及其对老年人护理的影响的知识。这需要跨多个分析尺度快速组织、分析和解释异构和噪声数据。最后，该项目正在开发新的技术，以支持参与老年人护理的所有行为者的不同和不可预测的信息需求：老年人本身、非正式护理人员（例如，家庭成员），正式的照顾者（例如，社区护士）和健康研究人员（例如，该项目的预期结果包括（a）集成各种异构家庭传感器的新平台，（B）解决家庭活动感测数据的噪声、低水平和突发性质的新流数据模型和查询处理技术，（c）用于与该数据进行有效终端用户交互的新技术，（d）将所有三个组成部分整合到一个单一的数据采集和分析管道中的系统，以及（e）深入了解这一变革性管道如何使社区老年护理受益。通过上述贡献，该项目将在美国的计算机科学和老年护理领域产生广泛的影响。我们的目标是以更低的成本提供更高质量的护理，让老年人在家里呆得更久。

Health care systems face critical challenges as people live longer with multiple chronic conditions that increase their need for health care services. Addressing the needs of an aging population requires improved methods for community-based preventive care through ongoing monitoring, early detection of adverse events and patterns, and early intervention.This project addresses this challenge by developing a transformative pipeline of flexible tools for at-home activity monitoring and analysis. The key goal is to enable at-home deployments of flexible sensor platforms and tools for agile analysis of this sensor data by users with varying skill sets and information needs. To achieve this goal, the project focuses on three specific challenges. First, this project is developing new platforms for deploying different combinations of at-home sensors according to varying elder needs. Second, this project is creating new tools for converting the raw sensor data into knowledge of elder activities and their impact on elder care. This requires rapid organization, analysis, and interpretation of heterogeneous and noisy data across multiple scales of analysis. Finally, the project is developing new techniques to support the varying and unpredictable information needs of all actors involved in elder care: the elders themselves, informal caregivers (e.g., family members), formal caregivers (e.g., community nurses), and health researchers (e.g., evaluating the effectiveness of different prescribed activity regimens).The expected results of this project include (a) new platforms that integrate a variety of heterogeneous home sensors, (b) new streaming data models and query processing techniques addressing the noisy, low-level, and bursty nature of home activity sensing data, (c) new techniques for effective end-user interaction with this data, (d) a system that integrates all three components into a single data acquisition and analysis pipeline, and (e) deep knowledge of how this transformative pipeline can benefit community elder care.Through the above contributions, this project will have broad impact in both computer science and elder care in the United States. The goal is to provide higher quality care at lower cost that keeps elders in their homes longer.