OBJECTIVE HOME MANAGEMENT OF PEDIATRIC ASTHMA EXACERBATION USING MOBILE TECHNOLOGY AND MACHINE LEARNING

项目总结 哮喘是美国最常见的慢性儿科疾病，影响620万人或1/4 12个孩子。尽管在儿童哮喘的管理方面取得了进展，但哮喘的恶化导致 约550,000急诊科(ED)就诊，80,000人住院，数百人 每年都会过早死亡。哮喘恶化的早期症状，特别是在儿童中， 无特异性，不幸的是，直到孩子出生后才被父母确认为哮喘 表现出更严重的症状，足以需要紧急护理。这样做的长期目标是 STTR倡议是通过补充以下内容使父母能够及时开始对急性哮喘进行治疗 他们的主观评估与客观衡量急性哮喘严重程度。移动设备 我们计划为此目标开发、测试和部署的技术将使用数字信号处理 (DSP)和机器学习(ML)来确定三个不同的严重程度区域(对应于 哮喘行动计划上的绿色、黄色和红色区域)允许父母遵循哮喘行动计划 准确地说。由此产生的改进和及时的以家庭为基础的儿童哮喘管理应该 减少目前过度使用ED和令人无法接受的高发病率和死亡率。这个 建议的技术受到儿科哮喘严重程度评分(PASS)的启发并基于此。通行证是 在许多儿科急诊室中用于客观评估急性哮喘的严重程度，以帮助管理 急性哮喘和严重的出院和住院决定。PASS由五个部分组成 经过充分研究和验证的临床参数。我们在提议的项目中的目标是开发新的 这项技术使父母能够在家庭环境中进行类似的测量并将其映射到 3个颜色编码的区域，便于执行哮喘行动计划。该项目的两个具体目标 是1)建立一个具有基本临床发现的儿童哮喘数据库，以及2)开发和 验证用于自动评估急性哮喘严重程度的DSP和ML算法。成功者 完成这些目标将产生一种经过验证的技术，适用于客观的、基于家庭的 评估急性哮喘的严重程度。这种评估目前是可能的，但仅限于医学上 由训练有素的医务人员提供设施。我们的第二阶段目标将是：(A)进一步改善和部署 (B)进行一项前瞻性试验，以衡量其可行性、利用率和 控制哮喘紧急情况和成本的有效性。在第二阶段，我们亦会推行 必要的监管批准。

PROJECT SUMMARY Asthma is the most common chronic pediatric disease in the United States, affecting 6.2 million or 1 in 12 children. Despite advances in the management of childhood asthma, asthma exacerbation results in approximately 550,000 emergency department (ED) visits, 80,000 hospitalizations, and hundreds of premature deaths each year. Early symptoms of asthma exacerbation, especially in children, are nonspecific and are unfortunately often not recognized by parents as asthma until the child demonstrates more severe symptoms, enough to require emergent care. The long-term goal of this STTR initiative is to empower parents to initiate timely therapy for acute asthma by supplementing their subjective assessment with an objective measure of acute asthma severity. The mobile technology we propose to develop, test and deploy toward this goal will use digital signal processing (DSP) and machine learning (ML) to determine three distinct severity zones (corresponding to the green, yellow and red zones on the asthma action plan) allowing parents to follow asthma action plans accurately. The resulting improved and timely home-based management of childhood asthma should reduce current excessive ED utilization and unacceptably high rates of morbidity and mortality. The proposed technology is inspired by and based on the pediatric asthma severity score (PASS). PASS is used in many pediatric EDs for objective assessment of acute asthma severity to aid management of acute asthma and critical discharge and hospitalization decisions. The components of PASS are five well-studied and validated clinical parameters. Our goal in the proposed project is to develop new technology that enables parents to make similar measurements in the home setting and map those to the 3 color-coded zones for easier execution of asthma action plans. The 2 specific aims of the project are to 1) build a pediatric asthma database with ground-truth clinical findings, and 2) develop and validate DSP and ML algorithms for automated assessment of acute asthma severity. The successful completion of these aims will result in a validated technology suitable for objective, home-based assessment of acute asthma severity. Such assessment is currently possible but only in medical facilities by trained medical staff. Our Phase II goals will be to (a) further improve and deploy the technology in homes and (b) conduct a prospective trial measuring its feasibility, utilization and effectiveness in controlling asthma emergencies and costs. In Phase II, we will also pursue the necessary regulatory approvals.