区域居民健康数据集

基于临床数据中心（CDR）架构，构建门诊就诊等临床数据模型矩阵搭建全域数据底座，抽取、转换、加载多源异构医疗数据。通过协议解析、归一化将非结构化、半结构化和结构化数据统一为标准格式，并通过重复数据检测、异常值检测算法，进行多级数据清洗。最终借助构建患者主索引，实现跨系统患者实体精准聚合，经业务逻辑建模，按时间轴和诊疗场景维度建模，通过规则引擎等数据质量校验方式，形成以患者主索引为核心的星型数据模型，为患者 360 视图提供数据支撑。Based on the Clinical Data Repository (CDR) architecture, this study constructs a clinical data model matrix covering scenarios including outpatient visits to establish a unified global data foundation. It leverages the Extract, Transform, Load (ETL) workflow to process multi-source heterogeneous medical data. Unstructured, semi-structured and structured data are standardized into a uniform format through protocol parsing and normalization, followed by multi-stage data cleaning with duplicate data detection and outlier detection algorithms. Ultimately, by building the Patient Master Index (PMI), accurate aggregation of patient entities across disparate systems is achieved. Following business logic modeling across dimensions of timeline and clinical diagnosis and treatment scenarios, and employing data quality validation approaches such as rule engines, a star-shaped data model with the Patient Master Index as the core is finally developed, which provides data support for the patient 360-degree view.

搜集汇总

背景与挑战

背景概述

该数据集由天津市东丽医院提供，包含272,620条居民健康记录，涵盖患者基本信息、就诊、检验、住院和体检等多维度临床数据，适用于医疗、教学和科研场景。数据通过临床数据中心架构进行清洗和建模，形成以患者主索引为核心的星型模型，支持区域健康分析和疾病防控策略制定，已进行电子存证确保数据安全。