传染病多维度分级预警研究

Infectious diseases are significant and emerging threats to public health, and may seriously affect socio-economic development and even national security of a country or a state. So the Chinese government is trying its best to minimize public health risks caused by infectious diseases.Infectious disease surveillance and early warning is significant to prevent and control infectious diseases.And to detect anomalies in the epidemiology of diseases earlier is one of the key technologies of infectious disease surveillance and early warning. However, existing methods are difficult to detect anomalies in various attribute dimensions of cases accurately. Most of those methods are retrospective studies.Therefore, the early warning of infectious diseases based on those methods can not send meaningful early warning signals.In addition, more and more surveillance resources are availabled. It requires a new early warning system with more powerful data compatibility. In order to deal with those problems, this project firstly detects the heterogeneity of a specific infectious disease in 'time-space-crowd' attribute demensions of cases. And then the project facilitates multi-dimensional trend analysis of the outbreak risk of the disease. Integrating the results of the above two models, the project proposes a new sub-level early warning format for infectious diseases. At last, the internet search query data from Baidu Index is used to assess the accuracy and timeliness of the results of the proposed early warning format. Those studies will provide new ideas to optimize the models of Chinese nationwide web-based automated system for early outbreak detection and rapid response. It may help to improve the government's capabilities of early warning of infectious diseases.

传染病暴发不仅直接影响人类健康，还会对社会经济发展，乃至国家安全产生严重影响。防控传染病的关键策略之一是传染病监测与预警，而尽早发现传染病异常发生与增加的苗头是实现这种策略的核心技术。但现有的预警技术很难精确探测到病例在不同维度上的异常，而且多为回顾性分析，并没有做到真正意义上的预警。另外，随着传染病监测渠道不断增加，也对现有预警系统的数据兼容功能提出了更高的需求。本项目针对以上问题，系统开展传染病多维度异常探测和暴发风险多维度趋势分析模型的研发，在此基础上提出一种全新的组件式传染病分级预警模式，并利用网络搜索数据完成预警模式结果的准确性和及时性的验证。研究成果将为我国传染病自动预警与响应系统平台概念模型的改进提供新的研究思路，为提高我国传染病预警防控能力提供基础。

传染病流行程度已经成为了一个国家或地区人口健康水准的重要指标之一。将传染病的危害控制在最低程度是国家或地区社会经济可持续性发展的重要保证，而且传染病预警能力也是国家或地区综合实力的重要体现。.本项目收集了湖南省各区县2009－2012 年手足口病监测数据（含病例和发病率数据）和同期与手足口病相关的关键字段网络搜索频率数据。应用智能算法和Bayesian 估计，得到了传染病在“时间-空间-人群”多维度上的异常聚集，从而对传染病进行多维度的异常探测。再基于空间面板模型的传染病暴发风险进行多维度趋势分析，建立组件式分级预警模式。最后利用同期网络搜索频率数据来论证所建预警模式的可行性、准确性和及时性。.本项目为我国传染病自动预警与响应系统平台概念模型的改进提供新的研究思路和方案，也为提高我国传染病预警防控能力提供坚实的理论和实践基础。