Moratilla Juan M, Alonso-Calvo Raul, Molina-Vaquero Gema, Paraiso-Medina Sergio, Perez-Rey David, Maojo Victor
Biomedical Informatics Group, DIA & DLSIIS, Facultad de Informática, Universidad Politécnica de Madrid, Campus de Montegancedo, Madrid, Spain.
Stud Health Technol Inform. 2013;192:971.
Breast cancer clinical trial researchers have to handle heterogeneous data coming from different data sources, overloading biomedical researchers when they need to query data for retrospective analysis. This paper presents the Common Data Model (CDM) proposed within the INTEGRATE EU project to homogenize data coming from different clinical partners. This CDM is based on the Reference Information Model (RIM) from the Health Level 7 (HL7) version 3. Semantic capabilities through an SPARQL endpoint were also required to ensure the sustainability of the platform. For the SPARQL endpoint implementation, a comparison has been carried out between a Relational SQL database + D2R and a RDF database. The results show that the first option can store all clinical data received from institutions participating in the project with a better performance. It has been also evaluated by the EU Commission within a patient recruitment demonstrator.
乳腺癌临床试验研究人员必须处理来自不同数据源的异构数据，这给生物医学研究人员在需要查询数据进行回顾性分析时带来了负担。本文介绍了在欧盟INTEGRATE项目中提出的通用数据模型（CDM），以统一来自不同临床合作伙伴的数据。该CDM基于健康级别7（HL7）第3版的参考信息模型（RIM）。还需要通过SPARQL端点提供语义功能，以确保平台的可持续性。对于SPARQL端点的实现，已经在关系SQL数据库+D2R和RDF数据库之间进行了比较。结果表明，第一种选择能够以更好的性能存储从参与项目的机构接收的所有临床数据。欧盟委员会还在一个患者招募示范项目中对其进行了评估。
