No abstract

Hospitals can supply clinical decision support systems with electronic healthcare record provided by different users and stored in different computer systems. This way, during a diagnosis decision a user can be supported by a remote knowledge database constructed by other users, like, for instance, a reference specialist discussion board. Usually, medical applications scenarios like this use different terminology systems leading to terminological interoperability problems. Still, clinical applications frequently make use of different clinical ontology systems that do not provide semantic types necessary to create a structure to carry shared knowledge database. The lack of this semantic types brings ontology interoperability problems which difficult the support of collaborative clinical decision-making. This paper presents a UMLS interoperable solution to promote clinical applications development intended to contribute with collaborative diagnosis decision, in a distributed computer environment based on service-oriented architecture. This paper still presents a case study to demonstrate DDSOnt usability through the development of a system prototype integrated to a collaborative and multimedia communication distance-learning environment.

Clinical Decision Support Systems (CDSS) are knowledge based systems that make use of patient data to generate a diagnosis response aiming to support a clinical decision. Usually, hospital and clinical systems supply CDSS with information generated by local users. Thus, CDSS's knowledge base may be created based on local experiences learning and will support users with this type of experience. A different and interesting situation would be if legacy systems could supply CDSS with clinical information deriving from physically remote systems, and created by physicians from others affiliations. At this scenario, a physician could be supported in a collaborative environment during a clinical decision with physicians working remotely, and with shared clinical knowledge bases. Concerning this distributed computer scenario technical interoperability, communication problems may appear mainly due to different communication protocol usage to create and invoke remote clinical services. Still considering the scenario presented before, semantic interoperability problems can occur because heterogeneous terminology and ontology clinical systems may be used by distinctive medical applications. Therefore, these interoperability problems must be explored and solved to allow a collaborative clinical decision support activities. This paper aims to investigate and discuss these technical and semantic interoperability problems caused by clinical information exchanged between CDSSs and health institutions systems. Based on that discussion, an interoperability and UMLS based solution is proposed to support a collaborative clinical decision environment on the Web. The solution is composed by a novel DDSOnt (Diagnosis Decision Support Ontology) ontology, that extends UMLS Semantic Network to create a new set of semantic types needful to define a structure that supports a shared knowledge databases to be used by physicians during the collaborative clinical decision. Ontology construction aims to promote ontology compatibility. Still, softwares have been developed to make DDSOnt useful by clinical applications that needs to be supported during clinical decision: a JDDSOnt API to define and to create semantic documents DDSOnt conformed, and a set of web services named DDSOntWs to enable a collaborative, service oriented, distributed computer environment, and plataform independent. Therefore, a CDSS is created as a web service. Aiming to promote terminology compatibility, one DDSOnt web service use the UMLS MetaThesaurus. An acute abdominal pain use case is also presented to demonstrate DDSOnt usability.

No abstract

No abstract