大数据驱动的智慧医疗健康管理创新

The proposed project aims to provide an effective response to the challenges of “inaccessibility and over-expensive provision of health and medical services”, and “emphasizing treatment rather than prevention”. Taking an interdisciplinary approach involving socioeconomic and healthcare informatics, the research addresses the key issues in health and medical service industry by providing decision support for health management and service innovation through effective mining and analysis of big data. The research involves fusion, integration and management of large healthcare data sets, with a focus on two major diseases of cardiovascular disease and tumour. The data possibly come from various novel sources such as the Internet of Things, social media, wearable devices, intelligent sense networks and other real-time data collection devices. Big data analytics to be deployed will involve data extraction, data storage, analysis and applications of real-time live healthcare sources. To further promote decision support for personalized medical monitoring and treatment for personal and home care, theoretical models, methods and techniques will be developed, such as the behavioral preference recognition model, personalized healthcare interference recommendation algorithms and clinical path dynamic programming methods. The research in the Internet of Things will lead to a set of factors in determining the medical service value and to developing medical service value models that can be applied to the current Chinese hierarchical clinic systems. The deliverables from the project will include the medical services evaluation index system for intelligent healthcare which will aid China's medical insurance reform, policy making at different levels and the New Rural Co-operative Medical System. The outcomes will be implemented and exploited through the empirical work by utilizing actual medical service management and decision making in public hospitals, with the aim of verifying, revising and enhancing the resultant theories and methods, in order to provide decision support through effective data management of intelligent healthcare and medical service innovation.

目前我国“看病贵、看病难”问题突出，“重治疗、轻预防”，大医院人满为患，疾病负担越来越重。本项目根据 “大数据驱动的管理与决策若干基础问题研究”重点项目群（G02）指南 (5) 的要求，基于国家人口健康科学数据共享平台数据资源，从经济管理学和医疗健康信息学角度，研究城乡医疗卫生管理与决策的相关问题。通过城乡医疗大数据集成，分析我国不同区域疾病负担，建立城镇医保和新农合卫生经济学模型；研究如何科学利用各类医疗卫生数据高效引导各类患者和人群真正实现分级诊疗的方法。研究尤其涵盖物联网、社交媒体和可穿戴设备等新数据源在智慧医疗和健康管理中的应用。基于电子病历和健康档案基础数据，开展医疗服务价值研究，建立重大慢性病最优诊疗路径知识库，为重点患者和人群提供个性化疾病诊疗和健康管理服务，并先在农村开展智慧医疗健康管理创新和示范。最终帮助提高我国医疗决策者管理水平和分级诊疗效率，解决看病贵和看病难的问题。

本项目属于“大数据驱动的管理与决策研究重大研究计划”中的重点项目群“大数据驱动的管理与决策若干基础问题研究”项目，主要关注大数据驱动的医疗健康管理与决策技术创新、社会服务范式创新等有关问题。执行期间每年参加基金委该重大研究计划的年度交流会。圆满完成任务书规定任务，主要取得如下三个方面的成果：（1）根据医疗大数据高维、多模的特点，研究医疗大数据语义和关联的挖掘与分析技术，提出智慧医疗服务价值相关理论和算法模型，面向患者和面向政府决策者两个层面，提出解决“看病贵、看病难”和“分级诊疗”中的突出问题的建议。提出基于医疗大数据分析进行临床误诊研判的新方法，使误诊学成为误诊科学成为可能。（2）搭建基于知识图谱的医疗大数据分析服务平台，提供系列结构化决策报告库和多种分析策略模型产生可视化的通用任务模型，智能生成多种决策分析报告，辅助用户进行决策；并提出通过该平台实现医疗大数据实时共享机制。（3）在安徽省阜南县和贵州省锦屏县等多个地区进行深度社会服务调研，以检验、修正和进一步发展大数据驱动智慧医疗的相关理论，算法，模型及服务平台，并服务地方取得显著经济效益和社会效益，提高区域医保资金使用效率和促进分级诊疗的精准实施。.具体成果如下：（1）以中国工程院的名义提交政策建议1项，获得时任副总理刘延东、马凯的批示，并抄送国家发改委、科技部、教育部等十部委参考落实；（2）共发表学术论文100篇，其中期刊论文76篇，会议论文24篇；（3）完成研究报告1项；出版学术专著1部、高等学校教材1部；（4）提交国家发明专利7项（已授权2项），登记软件著作权2项；（5）培养博士生12人，硕士生56人，在大数据驱动的医疗健康管理领域，造就了一支具有国际竞争力的研究队伍；（6）举办国际学术会议5次；（7）建立大数据分析平台1套；（8）获得包括国家自科重点基金项目、国家社会科学基金重大项目在内的新立相关项目多项。