SCH: Wearables for Health and Disease Knowledge (W4H)

项目说明 1简介 与近代史上的任何其他现象相比，新冠肺炎大流行对 我们处理病人护理是因为它给医院、诊所和健康带来了巨大的负担 专业人士。卫生界已经通过研究和技术对这一趋势做出了回应 利用超出通常所认为的“健康数据”的数据，例如通信- 身份和背景数据、社交媒体、流量和移动性数据。例如，Nsoesie等人[84] 分析武汉的医院交通和搜索引擎数据，推断秋季早期疾病活动 2019年。这些新的努力，包括我们自己利用移动性数据预测COVID的工作- 19‘S传播风险[94]，使用了美国国家科学基金会的这项号召提案所指的“非传统健康” 数据“。 在这份提案中，我们重点关注一种特定类型的非传统健康数据，即可穿戴数据，它 也迅速成为健康和疾病数据的重要来源，因为它们提供了各种信息 与个人、行为、社会、背景和环境健康相关的因素。可穿戴设备 主要用于活动跟踪[96，15，20，80]，并随着健身而流行 应用；然而，最近，这些设备被用于越来越多的 健康应用，包括健康监测、临床护理、远程临床试验、药物输送、 以及疾病的特征等等。事实上，可穿戴设备在许多方面都很有用- 申请和疾病(例如帕金森氏病、癫痫和中风[57]、睡眠障碍-- DES[12]、心脏疾病[90、63]和癌症[75])。随着新冠肺炎的推进，这一趋势正在加速 流行病，比如智能手机已经被提出用来跟踪症状，监测有效性[] 非药物干预措施，评估潜在传播，并支持接触者追踪[45]。 可穿戴测量不同于传统的临床测量。当病人到医院就诊时 临床、生命体征和实验室检查是在“受控”的环境中短时间内收集的，使用 多台设备。我们将受控环境中的这种监视定义为诊所内快照 监控，缩写为SIC。与此同时，可穿戴设备最近的增长和可及性 带有嵌入式活动和移动传感器的智能手机和手表[97]等设备[114] 能够长时间持续监测患者的生命体征和其他健康指标 持续时间。使用可穿戴设备进行患者监控通常是在一种“不受控制”的情况下进行的 在家中或在工作中以非侵入性方式设置，只需几个传感器。这一趋势也已经 被NIH的mHealth倡议所概括，导致了新的医疗保健的发展 “家庭医疗”[9，40]和“分钟诊所”[125]等模式，与 无论是智能手机中无处不在的传感器，还是血糖监测仪等定制传感器[62]。我们定义 这种在不受控制的环境中的监控称为纵向现场监控，简称 就像生命一样。显然，这些都是文字游戏，也就是说，SIC是用来捕捉病人在生病时的精神状态的 当病人在家中和工作中过着正常的“生活”时，他们会去诊所/医院，而不是生活。 生命监护占据了患者99%以上的时间，实现了门诊监护 疾病及其治疗对患者表现和生活质量的影响。事实上，我们的 初步数据显示，在某些情况下，例如评估癌症的表现状态 患者，生活数据优于办公室SIC评估[82]。 SIC监测是当前的护理标准，并由改善可测量的结果推动 第72页

Project Description 1 Introduction More than any other phenomena in recent history, the COVID-19 pandemic has challenged how we approach patient-care due to the huge burden it has placed on hospitals, clinics, and health professionals. The health community has responded to this trend with research and technology leveraging data that goes beyond what is customarily thought of as “health data”, such as commu- nity and contextual data, social media, traffic, and mobility data. For example, Nsoesie et al.[84] analyzed hospital traffic and search engine data in Wuhan to infer early disease activity in Fall 2019. These new efforts, including our own work in utilizing mobility data to forecast COVID- 19’s transmission risk [94], uses what this NSF call-for-proposal refers to as “non-traditional health data”. In this proposal, we focus on one specific type of non-traditional health data, wearable data, which are also fast becoming an important source of health and disease data as they inform on a variety of personal, behavioral, social, contextual, and environmental health-relevant factors. Wearables have been primarily used for activity tracking [96, 15, 20, 80] and gained popularity with fitness applications; however, more recently, these devices have been used in an increasing number of health applications, including health monitoring, clinical-care, remote clinical-trials, drug delivery, and disease characterization to name a few. In fact, wearables have been found useful in a num- ber of applications and diseases (e.g., Parkinson’s disease, epilepsy and stroke [57], sleep disor- ders [12], cardiac disorders [90, 63] and cancer [75]). This trend is accelerating with the COVID-19 epidemic, e.g., smartphones have been proposed to track symptoms [64], monitor effectiveness of non-pharmaceutical interventions, assess potential spread, and support contact tracing [45]. Wearable measurements differ from traditional clinical measurements. When a patient visits a clinic, vitals and lab tests are collected in a “controlled” environment in a short duration of time using multiple devices. We define this monitoring in the controlled environment as Snapshot In-Clinic monitoring, abbreviated as SIC. Meanwhile, the recent growth and accessibility of the wearable devices such as smartphones and watches [97] with embedded activity and mobile sensors [114] enables the continuous monitoring of patients’ vital signs and other health indicators over a long duration of time. Patient monitoring using wearable devices typically happens in an “uncontrolled” setup at home or at work in a non-intrusive fashion with only a few sensors. This trend has also been encapsulated by the NIH mHealth’s initiatives, resulting in the evolution of new healthcare models such as “home healthcare” [9, 40] and “minute clinic” [125], which goes hand in hand with both ubiquitous sensors in smartphones and custom sensors like glucose monitors [62]. We define this monitoring in the uncontrolled environment as Longitudinal In-Field monitoring, abbreviated as LIFE. Clearly these are wordplay, i.e., SIC is for “sick” capturing patients’ state of mind when they visit a clinic/hospital vs. LIFE for when patients live their normal “life” at home and at work. LIFE monitoring makes up for greater than 99% of patients’ time, enabling outpatient monitoring of the effects of disease and its therapy on patient performance and quality of life. In fact, our preliminary data show that in some cases, such as assessment of performance status in cancer patients, LIFE data outperform in-office SIC assessments [82]. SIC monitoring is the current standard of care and is driven by improving outcomes in measurable Page 72