Topic 412: CHAMP-WARE: Continuous Health Monitoring and Predictions using a Wearables-Agnostics Platform for Cancer Patients, Phase I

主题 412：CHAMP-WARE：使用可穿戴设备不可知论平台对癌症患者进行持续健康监测和预测，第一阶段

• 批准号: 10265744
• 负责人: Sridhar Ramakrishnan
• 金额: $ 40万
• 依托单位: INTELLIGENT AUTOMATION, INC.
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-09-16 至 2021-06-15
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10265744
• 关键词: Adverse event; Algorithmic Analysis; Artificial Intelligence; Biological Markers; Cancer Patient; Cancer Remission; Clinic; Clinical; Clinical Trials Design; Collaborations; Computer software; Data; Data Aggregation; Data Analyses; Data Analytics; Data Set; Data Storage and Retrieval; Databases; Detection; Devices; Documentation; Future; Health; Health Care Costs; Infrastructure; Intelligence; Internet; Machine Learning; Mainstreaming; Measurement; Measures; Monitor; Oncology; Outcome; Patient Monitoring; Patients; Phase; Physiological; Physiology; Population Research; Predictive Analytics; Quality of life; Reporting; Research; Research Personnel; Sampling; Secure; Signal Transduction; Source; Standardization; Stream; Symptoms; System; Therapeutic; Time; Vendor; Work; anticancer research; cancer care; clinically relevant; cohort; data standards; data streams; data visualization; digital; flexibility; functional outcomes; improved; informatics tool; population based; prototype; repository; response; sensor; side effect; smartphone Application; software systems; standard of care; statistics; tool; tumor progression; web portal

Commercially available off-the-shelves (COTS) wearables that objectively track physiological variables offer a rich source of information about a patient’s health to clinicians and oncology researchers, to facilitate early adverse-event detection and subsequent management, which can decrease healthcare costs and improve patient quality of life. The passive, continuously measured data streams generated by current or future COTS sensors will allow direct/indirect measures of cancer progression and its symptoms. Increased out-of-clinic patient and clinician engagement via these tools will allow more precise delivery of cancer care after as well as during cancer remission. Ultimately, these passive sensing platforms’ data for digital biomarkers will afford clinicians: 1) more objective metrics of response to therapeutics; 2) control and autoreporting of symptoms and their fluctuations; 3) monitoring of side-effects of experimental or standard of care therapies; and 4) more ecologically valid clinical endpoints, all decreasing assessment burden via increased continuity of physiological measurement sampling and patient context in the ambulatory setting. Furthermore, such data present an opportunity to measure population-based statistics from large cohorts of cancer patients by way of the myriad of devices currently available or being developed. Unfortunately, despite the availability of a myriad of COTS wearables capable of measuring physiological variables, their use for remote cancer patient monitoring or for out-of-clinic cancer research is yet to become mainstream. There is a considerable need for scalable informatics tools that allow automated data aggregation, integration and machine learning/artificial intelligence (AI)/predictive analytics that can pull from disparate data sets across COTS device vendors and have the flexibility to add new measures as they are developed. Furthermore, a central software platform is needed that could obtain wearable or external device data and uniformly compare/contrast/couple data streams to understand physiology versus patient context with respect to time: such a capability will substantially advance this unique approach to aid cancer patients, clinician assessment and clinical trial design. This work seeks to overcome these bottlenecks and provide a workflow and an infrastructure for out-of-clinic remote patient monitoring and online research collaboration for advancing population-based research. By developing a software system, comprised of a smartphone app, database, and a Web portal, which can a) collect and standardize raw sensor data from multitude of wearables, b) perform intelligent multi-sensor data analytics to provide clinically relevant outcomes in real time, c) store these data in a common repository, and d) provide online interfaces to view and analyze data, the proposed effort will significantly advance out-of-clinic cancer research and patient monitoring.

商用现货（COTS）可穿戴设备， 生理变量提供了关于患者健康的丰富信息源， 临床医生和肿瘤学研究人员，以促进早期不良事件检测， 后续管理，可降低医疗成本并提高患者质量 生命由当前或未来的数据流生成的被动的、连续测量的数据流 COTS传感器将允许直接/间接测量癌症进展及其 症状通过这些工具提高门诊患者和临床医生的参与度， 允许在癌症缓解后以及癌症缓解期间更精确地提供癌症护理。 最终，这些被动传感平台的数字生物标志物数据将提供 临床医生：1）更客观的治疗反应指标; 2）控制和自动报告 症状及其波动; 3）监测的副作用 实验性或标准护理疗法;以及4）更生态有效的临床 终点，所有通过增加生理连续性降低评估负担 测量采样和门诊环境中的患者背景。此外，委员会认为， 这些数据提供了一个机会， 通过目前可用或正在使用的无数设备， 开发 不幸的是，尽管有无数的COTS可穿戴设备， 测量生理变量，它们用于远程癌症患者监测或用于 临床外癌症研究尚未成为主流。有相当大的需要 用于可扩展的信息学工具，这些工具允许自动化数据聚合、集成和 机器学习/人工智能（AI）/预测分析，可以从 跨COTS设备供应商的不同数据集，并具有灵活性， 制定的措施。此外，还需要一个中央软件平台， 可以获得可穿戴或外部设备数据，并统一比较/对比/耦合 数据流，以了解生理学与患者背景相对于时间的关系：例如， 能力将大大推进这种独特的方法，以帮助癌症患者，临床医生 评估和临床试验设计。 这项工作旨在克服这些瓶颈，并提供一个工作流程和一个 用于门诊外远程患者监测和在线研究的基础设施 合作推进基于人口的研究。通过开发一个软件 系统，由智能手机应用程序、数据库和Web门户组成，可以a） 收集和标准化来自大量可穿戴设备的原始传感器数据，B）执行 智能多传感器数据分析，以提供真实的临床相关结果 时间，c）将这些数据存储在公共存储库中，以及d）提供在线界面以查看 并分析数据，这项计划将大大推进临床外癌症的治疗。 研究和患者监测。