多源数据在流感预测中的应用研究

阎文蓉; 宋   阳; 陈邦华; 马敬东

文章摘要

阎文蓉,宋阳,陈邦华,马敬东.多源数据在流感预测中的应用研究[J].中华医学图书情报杂志,2022,31(9):12-19.

多源数据在流感预测中的应用研究

Application of multi-source data in influenza prediction

投稿时间：2022-08-10

DOI：10.3969/j.issn.1671-3982.2022.09.002

中文关键词: 流感预测预警监测机器学习

英文关键词: Influenza prediction Early warning Surveillance Machine learning

基金项目:武汉市预防医学科研专项重大项目“突发公共卫生事件应急决策支持系统和应对处置能力建设研究”（WY22M03）

作者	单位
阎文蓉	华中科技大学同济医学院医药卫生管理学院，湖北武汉 430030
宋阳	华中科技大学同济医学院医药卫生管理学院，湖北武汉 430030
陈邦华	武汉市疾病预防控制中心，湖北武汉 430022
马敬东	华中科技大学同济医学院医药卫生管理学院，湖北武汉 430030

摘要点击次数: 76

全文下载次数: 39

中文摘要:

目的：梳理目前流感预测研究中多源数据的应用，为流感等传染病监测研究提供启示和参考。方法：系统检索Web of Science数据库、Scopus数据库、PubMed数据库、中国知网（CNKI）数据库、万方数据知识服务平台和维普中文期刊服务平台，检索时间为2009年1月至2022年4月。纳入基于多源数据进行流感预测的相关研究文献，并对符合纳入和排除标准的文献进行评述。结果：共纳入文献115篇，根据多源数据使用频率，从高至低依次为互联网数据、环境数据、症状监测数据和组合数据，基于多源数据的流感预测模型包括传染病模型、时空模型、机器学习模型和集成模型。结论：基于传统监测系统和新型数据源的预测模型在提升流感预测预警能力上已取得成效，但仍需从数据质量提升、预测模型优化、预警系统平台和工具改进、信息技术集成等方面进一步提高预测的准确性和及时性。

英文摘要:

Objective The application studies of multi-source data in current influenza prediction were sorted out to provide insights and reference for influenza and other infectious disease surveillance studies. Methods Literature in Web of Science, Scopus, PubMed, Chinese National Knowledge Infrastructure (CNKI), Wanfang Data Knowledge Service Platform and VIP Database for Chinese Technical Periodicals was systematically searched from January 2009 to April 2022. Literature on studies related to influenza prediction using multi-source of data was included, and literature that met the inclusion and exclusion criteria was reviewed. Results A total of 115 papers were included and were internet data, environmental data, symptom surveillance data, and combined data according to the frequency of multi-source data use in descending order. Influenza prediction models based on multi-source data included infectious disease models, spatio-temporal models, machine learning models, and integrated models. Conclusion Prediction models based on traditional surveillance systems and new data sources have been effective in improving influenza prediction and early warning capabilities, but further improvements in prediction accuracy and timeliness are still needed in terms of data quality improvement, prediction model optimization, early warning system platform and tool improvement, and information technology integration.

HTML 查看全文查看/发表评论下载PDF阅读器