Marco-Ruiz Luis, Pedrinaci Carlos, Maldonado J A, Panziera Luca, Chen Rong, Bellika J Gustav
Norwegian Centre for e-Health Research, University Hospital of North Norway, P.O. Box 35, N-9038 Tromsø, Norway; Department of Clinical Medicine, Faculty of Health Sciences, UIT The Arctic University of Norway, 9037 Tromsø, Norway.
Knowledge Media Institute, The Open University Walton Hall, Milton Keynes MK7 6AA, UK.
J Biomed Inform. 2016 Aug;62:243-64. doi: 10.1016/j.jbi.2016.07.011. Epub 2016 Jul 8.
The high costs involved in the development of Clinical Decision Support Systems (CDSS) make it necessary to share their functionality across different systems and organizations. Service Oriented Architectures (SOA) have been proposed to allow reusing CDSS by encapsulating them in a Web service. However, strong barriers in sharing CDS functionality are still present as a consequence of lack of expressiveness of services' interfaces. Linked Services are the evolution of the Semantic Web Services paradigm to process Linked Data. They aim to provide semantic descriptions over SOA implementations to overcome the limitations derived from the syntactic nature of Web services technologies.
To facilitate the publication, discovery and interoperability of CDS services by evolving them into Linked Services that expose their interfaces as Linked Data.
We developed methods and models to enhance CDS SOA as Linked Services that define a rich semantic layer based on machine interpretable ontologies that powers their interoperability and reuse. These ontologies provided unambiguous descriptions of CDS services properties to expose them to the Web of Data.
We developed models compliant with Linked Data principles to create a semantic representation of the components that compose CDS services. To evaluate our approach we implemented a set of CDS Linked Services using a Web service definition ontology. The definitions of Web services were linked to the models developed in order to attach unambiguous semantics to the service components. All models were bound to SNOMED-CT and public ontologies (e.g. Dublin Core) in order to count on a lingua franca to explore them. Discovery and analysis of CDS services based on machine interpretable models was performed reasoning over the ontologies built.
Linked Services can be used effectively to expose CDS services to the Web of Data by building on current CDS standards. This allows building shared Linked Knowledge Bases to provide machine interpretable semantics to the CDS service description alleviating the challenges on interoperability and reuse. Linked Services allow for building 'digital libraries' of distributed CDS services that can be hosted and maintained in different organizations.
临床决策支持系统（CDSS）开发成本高昂，因此有必要在不同系统和组织之间共享其功能。有人提出采用面向服务的架构（SOA），通过将CDSS封装在Web服务中来实现其重用。然而，由于服务接口缺乏表现力，CDS功能共享仍存在巨大障碍。链接服务是语义Web服务范式的演进，用于处理链接数据。它们旨在为SOA实现提供语义描述，以克服Web服务技术语法性质带来的限制。
通过将CDS服务演化为以链接数据形式公开其接口的链接服务，促进CDS服务的发布、发现和互操作性。
我们开发了方法和模型，将CDS SOA增强为链接服务，基于机器可解释的本体定义丰富的语义层，以支持其互操作性和重用。这些本体提供了CDS服务属性的明确描述，以便将其公开到数据网络。
我们开发了符合链接数据原则的模型，以创建构成CDS服务的组件的语义表示。为了评估我们的方法，我们使用Web服务定义本体实现了一组CDS链接服务。Web服务的定义与所开发的模型相关联，以便为服务组件附加明确的语义。所有模型都与SNOMED-CT和公共本体（如都柏林核心）绑定，以便依靠一种通用语言来探索它们。基于机器可解释模型对CDS服务进行发现和分析，并对所构建的本体进行推理。
通过基于当前的CDS标准，链接服务可以有效地将CDS服务公开到数据网络。这允许构建共享的链接知识库，为CDS服务描述提供机器可解释的语义，减轻互操作性和重用方面的挑战。链接服务允许构建分布式CDS服务的“数字图书馆”，这些服务可以在不同组织中托管和维护。
