IHBEM: Empirical analysis of a data-driven multiscale metapopulation mobility network modeling infection dynamics and mobility responses in rural States

农村流行病学是流行病学的一个重要分支，旨在研究农村社区的特征如何影响健康。虽然不到五分之一的美国人口生活在农村地区，但这些地区占美国土地总面积的97%左右，因此农村地区的人口密度很低。传统上，主流流行病学模型关注人口密度高的环境，如美国的主要城市。为了研究农村流行病学，有必要开发适合农村地区独特的人口统计和地理隔离的流行病学模型和分析。这项拨款汇集了数学建模、高性能计算、地理信息系统和医学社会学方面的专业知识，为怀俄明州、蒙大拿州、爱达荷州和科罗拉多州等西部山区各州开发可靠、准确的实时流行病预测框架。该项目涉及构建一个数据驱动的多尺度移动网络，结合使用无偏见移动电话数据、美国社区调查（ACS）和这些地区的人口普查数据。通过多尺度流动网络，主要研究人员将能够研究大流行期间的遏制政策（如减少流动干预措施、居家令和国家边境封锁）如何影响传染病的传播动态，以及此类政策如何加剧农村地区公共卫生差距和社会不平等。这项研究将有助于政府卫生官员在未来的流行病中为农村地区制定公平的政策和缓解战略。此外，拟议的社区外展将提高怀俄明州对疾病预防的认识。这项资助为农村流行病学的三个研究问题提供了解决方案：如何构建适合农村地区的超人口流动网络？人类的流动性是推动传染病在广泛地理范围内传播的因素之一。该项目在多尺度数据驱动的流动网络中引入了一个元人口模型，该模型是根据农村人口密度的独特结构和人类移动模式量身定制的，用于模拟前沿各州的感染动态。对于这种元人口流动模型，pi将适应并提高Gillespie事件驱动算法的效率，并推导出适合大规模随机模拟的合适时间步长。pi还将实现一种有效的粒子滤波算法，用于数据同化，以估计两个尺度上的局部传输参数。Q2。如何检测和纠正手机数据偏差？如果没有解决用于构建移动网络的移动电话数据的偏差，感染动力学模拟可能会错误地估计疾病传播的严重程度和速度。该项目将调查用于构建移动网络的移动电话数据的偏差，并开发一种方法来抵消数据偏差。第三季。遏制措施对弱势种族和社会经济群体有什么影响？pi将使用提议的无偏多尺度超人口流动模型来研究暂时减少流动性的影响及其对前沿国家中处于不利地位的种族和社会经济群体的持久行为改变效应。该项目由MPS数学科学部（DMS）通过数学生物学计划，SBE社会和经济科学部（SES）以及刺激竞争研究的既定计划（EPSCoR）共同资助。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Rural epidemiology is a critical subfield of epidemiology seeking to examine how the characteristics of rural communities influence health. Although less than one-fifth of the U.S. population lives in rural areas, these areas encompass about 97% of the total land area in the United States and thus the population density for rural areas is low. Traditionally, mainstream epidemiological models focus on settings with high population density, such as major U.S. cities. To study rural epidemiology, it is necessary to develop epidemiological models and analyses tailored to the unique demographics and geographic isolation of rural areas. This grant brings together expertise in mathematical modeling, high-performance computing, geographic information systems, and medical sociology to develop a framework for reliable and accurate real-time epidemic forecasts for mountain-West States, such as Wyoming, Montana, Idaho, and Colorado. The project involves constructing a data-driven multiscale mobility network, using the combination of de-biased mobile phone data, American Community Survey (ACS), and census data in these regions. With a multiscale mobility network, the principal investigators (PIs) will be able to examine how containment policies during a pandemic, such as mobility reduction interventions, stay-at-home orders, and State border lockdowns, impact the transmission dynamics of infectious diseases, and how such policies may exacerbate the disparities of public health and social inequalities in rural areas. The study will help government health officials develop equitable policies and mitigation strategies for rural places in the future epidemic. Furthermore, the proposed community outreach will raise awareness of disease prevention in Wyoming. This grant offers solutions to three research questions in rural epidemiology: Q1. How to construct a metapopulation mobility network tailored to rural places? Human mobility is one of the factors driving the spread of infectious diseases across wide geographical ranges. The project introduces a metapopulation model embedded in a multiscale data-driven mobility network tailored to the unique structures of rural population density and human moving patterns for simulating infection dynamics of the front-range States. For this metapopulation mobility model, the PIs will adapt and improve the efficiency of Gillespie’s event-driven algorithms and derive a proper time step for large-scale stochastic simulations. The PIs will also implement an efficient particle filter algorithm for data assimilations to estimate local transmission parameters at two scales. Q2. How to detect and correct mobile phone data bias? If the bias of the mobile phone data used to construct the mobility networks is not addressed, infection dynamics simulations may incorrectly estimate the severity and speed of disease transmission. The project will investigate the bias of mobile phone data used to construct the mobility network and develop a method to offset the data bias. Q3. What are the impacts of containment measures on disadvantaged racial and socioeconomic groups? The PIs will use the proposed unbiased multiscale metapopulation mobility model to study the impacts of temporary mobility reductions and their lasting behavior-modifying effects on the disadvantaged racial and socioeconomic groups in the front-range States.This project is jointly funded by the MPS Division of Mathematical Sciences (DMS) through the Mathematical Biology Program, the SBE Division of Social and Economic Sciences (SES), and the Established Program to Stimulate Competitive Research (EPSCoR).This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.