Mori A R, Galeazzi E, Consorti F, Bidgood W D
Reparto Informatica Medica, ITBM-CNR, Roma, Italy.
Proc AMIA Annu Fall Symp. 1997:650-4.
We developed a technique of reverse engineering to extract a conceptual schema--also called "categorial structure" in the European standard CEN ENV 12264 (MoSe)--from a set of terminological phrases. The technique was originally applied to coding systems, ie. to large value sets. We applied this technique to subsets of two new terminological resources for message standards: headings of patient record from Clinical LOINC and names of "Context Groups" for structured reporting from SNOMED DICOM Microglossary (SDM). Both sources provide context-independent names for message fields and domains of admitted values. Therefore conceptual schemata on the potential content for a field are compatible with the ones on names of the fields themselves. Both kinds of schemata can be compared and integrated with conceptual schemata for the information system that manages the patient record. This continuity in the schemata allows the coupling of applications with different organization of data, and will facilitate mapping from an application to standard messages and viceversa. Moreover, the simplified representation produced according the MoSe approach is easy to understand by healthcare operators, allowing their progressive involvement in cooperative efforts of design, discussion and validation of the schemata.
我们开发了一种逆向工程技术,用于从一组术语短语中提取概念模式——在欧洲标准CEN ENV 12264(MoSe)中也称为“范畴结构”。该技术最初应用于编码系统,即大型值集。我们将此技术应用于两种新的消息标准术语资源的子集:Clinical LOINC的患者记录标题和SNOMED DICOM微词汇表(SDM)结构化报告的“上下文组”名称。这两个来源都为消息字段和允许值的域提供了与上下文无关的名称。因此,关于字段潜在内容的概念模式与字段本身名称的概念模式是兼容的。这两种模式都可以与管理患者记录的信息系统的概念模式进行比较和整合。模式中的这种连续性允许不同数据组织的应用程序进行耦合,并将便于从应用程序到标准消息的映射,反之亦然。此外,根据MoSe方法生成的简化表示易于医疗保健操作人员理解,使他们能够逐步参与模式设计、讨论和验证的合作工作。