Communicable diseases, such as influenza, are transmitted from individual to individual following a network of contacts in a population. Reports on recent outbreaks, such as SARS, the Bird flu, and the H1N1 flu, have repeatedly stressed the critical role of contact networks. We propose an innovative Three-population and Three-scale Social Network (3p3sNet) model to simulating the spatial and temporal dispersion of influenza in a metropolitan population in Buffalo, NY. The 3p3sNet aims to construct a realistic contact network by representing interacting and mobile behaviors of individuals at three scales and three types of places. These involve individual (microscopic) -> local network (mesoscopic) -> population (macroscopic) as nighttime population at homes, daytime population at workplaces, and pastime population at service places. Through this network, diseases disperse from infectious individuals to their local networks then to the population-wide network in a complex dynamic fashion. Modeling the disease dispersion through this network provides invaluable insights in who might be at risk, where and when this risk might occur, and with whom these at-risk individuals might be in contact. These insights lay the foundation of developing spatially and temporally sensitive intervention strategies targeted towards the most vulnerable individuals and social groups. Furthermore, the 3p3sNet can be applied in modeling the epidemiology of any disease where human contacts play a critical role. In implementing 3p3sNet, we propose to use mobile phone data to extract the individual interaction and travel behaviors. We embrace recent developments in economics, geostatistics, econometrics, and machine learning to construct the network. We develop an innovative co-kriging approach to expanding local households to population-wide households and a novel distance-based GEV discrete choice model to link homes to workplaces and service places. It is anticipated that the assemblage of these advanced methods will enable new capabilities and bring transformative improvements in health-related studies in metropolitan areas. We will conduct a data-rich validation process for the three constructed populations, the links between them, and the simulated disease dispersion through the population. A comprehensive range of independent datasets will be used to support the proposed validation. These involve high-resolution population, workplace, and service place data, surveys of individual interaction and travel behavior, and reports on influenza infections. The multidisciplinary team comprises world-renown leaders and scholars in epidemiology, agent-based and social network modeling, human mobility analysis, geographical information science, and machine learning. The proposed project represents emerging frontiers in the modeling of communicable diseases and will redefine the capabilities of epidemiological models.
This project addresses two issues of central importance to successfully capturing the complex, spatial and temporal dispersion process of a communicable disease through a population. They are: how to represent individuals as heterogeneous, mobile, and interacting and how to model the disease dispersion process from infectious individuals to their local networks then to the population-wide network. To address these issues, this project proposes to: (1) use mobile phone data to construct a three-population and three-scale social network (3p3sNet); (2) simulate the spatially and temporally dynamic dispersion of influenza through an urban population in Buffalo, NY; and (3) conduct an intensive, data-rich validation process on the simulated influenza dispersion.
| Guan, Xiangyang; Chen, Cynthia (2018) General methodology for inferring failure-spreading dynamics in networks. Proc Natl Acad Sci U S A 115:E8125-E8134 |
| Wang, Feilong; Chen, Cynthia (2018) On data processing required to derive mobility patterns from passively-generated mobile phone data. Transp Res Part C Emerg Technol 87:58-74 |
| Mennis, Jeremy; Yoo, Eun-Hye Enki (2018) Geographic Information Science and the Analysis of Place and Health. Trans GIS 22:842-854 |
| Gao, Peng; Bian, Ling (2016) Scale Effects on Spatially Embedded Contact Networks. Comput Environ Urban Syst 59:142-151 |
| Zhong, Shiran; Bian, Ling (2016) A location-centric network approach to analyzing epidemic dynamics. Ann Am Assoc Geogr 106:480-488 |
| Chen, Cynthia; Ma, Jingtao; Susilo, Yusak et al. (2016) The promises of big data and small data for travel behavior (aka human mobility) analysis. Transp Res Part C Emerg Technol 68:285-299 |
