Search

As Electronic Healthcare Records become more prevalent, there is an increasing need to ensure unambiguous data capture, interpretation, and exchange within and across heterogeneous applications. To address this need, a common, uniform, and comprehensive approach for representing clinical information is essential. At Partners HealthCare System, we are investigating the development and implementation of enterprise-wide information models to specify the representation of clinical information to support semantic interoperability. This paper summarizes our early experiences in: (1) defining a process for information model development, (2) reviewing and comparing existing healthcare information models, (3) identifying requirements for representation of laboratory and clinical observations, and (4) exploring linkages to existing terminology and data standards. These initial findings provide insight to the various challenges ahead and guidance on next steps for adoption of information models at our organization.

Structured Clinical Documentation is a fundamental component of the healthcare enterprise, linking both clinical (e.g., electronic health record, clinical decision support) and administrative functions (e.g., evaluation and management coding, billing). One of the challenges in creating good quality documentation templates has been the inability to address specialized clinical disciplines and adapt to local clinical practices. A one-size-fits-all approach leads to poor adoption and inefficiencies in the documentation process. On the other hand, the cost associated with manual generation of documentation templates is significant. Consequently there is a need for at least partial automation of the template generation process. We propose an approach and methodology for the creation of structured documentation templates for diabetes using Natural Language Processing (NLP).

The design and construction of domain specific ontologies and taxonomies requires allocation of huge resources in terms of cost and time. These efforts are human intensive and we need to explore ways of minimizing human involvement and other resources. In the biomedical domain, we seek to leverage resources such as the UMLS Metathesaurus and NLP-based applications such as MetaMap in conjunction with statistical clustering techniques, to (partially) automate the process. This is expected to be useful to the team involved in developing MeSH and other biomedical taxonomies to identify gaps in the existing taxonomies, and to be able to quickly bootstrap taxonomy generation for new research areas in biomedical informatics.

Information overload, arising from di erent types of heterogeneous digital data readily accessible from millions of repositories, is a critical problem on the Global Information Infrastructure (GII). We present an information brokering approach, architecture and techniques that address issues related to information overload on the GII. The approach spans three levels: representation (structure/format/type) of digital data, information content captured in the data; and the vocabulary underlying the data. Metadata (data/information about data) is used to abstract from heterogeneous representational details and capture information content. Domain speci c ontologies are used to represent and interoperate across di erent vocabularies used to characterize information content. The approach thus suggested induces a metadata-based architecture that enables information brokering at the di erent levels. The feasibility of the approach is demonstrated by using a wide variety of metadata to capture information content for textual, image and structured data. These metadata belong to a wide spectrum and may range from metadata independent of the data content to those capturing information content in a application and domain speci c manner. This thesis demonstrates how metadata characterizing information in a domain speci c manner may enable: (a) media-independent correlation of information across ii heterogeneous media; and (b) vocabulary-based interoperation of information across di erent domains. Example information brokering prototypes based on metadata capturing information content to varying degrees are presented as instantiations to validate the proposed architecture. We also identify the desired (SEA") properties of an architecture in the presence of information overload, namely, scalability, extensibility and adaptability; and discuss in what measure the prototypes display these properties. The intrinsic trade-o between scalability and extensibility is identi ed and discussed. Adaptability, a new proposed property, is the ability of an information brokering system to adapt to di erent vocabularies used to describe similar information content. We show how maximizing scalability leads to issues of adaptability and how terminological relationships across domain speci c ontologies characterizing vocabularies may be used to achieve interoperation and increase adaptability.

The rapid advances in computer and communication technologies, and their merger, is leading to a global information market place. It will consist of federations of very large number of information systems that will cooperate to varying extents to support the users' information needs. We propose an architecture which may facilitate meeting these needs. It consists of three main components: information providers, information brokers and information consumers. We also propose an approach to information brokering. We discuss two of it's tasks: information resource discovery, which identities relevant information sources for a given query, and query processing, which involves the generation of appropriate mapping from relevant but structurally heterogeneous objects. Query processing consists of information focusing and information correlation. While the access-based search, and syntactic and hierarchical information organization has been adequate in the past, information brokering in presence of huge digital libraries or millions of information sources will likely require semantics and information-content based search and structuring of information. Our approach is based on: semantic proximity, which represents semantic similarities based on the context of comparison, and schema correspondences which are used to represent structural mappings and are associated with the context. The context of comparison of the two objects is the primary vehicle to represent the semantics for determining semantic proximity. Specifically, we use a context to capture the semantics in terms of the meaning and/or the use of an object. Using a partial context representation, we capture the assumptions in the intended use of the objects and the intended meaning of the user query. Information focusing is supported by subsequent context comparison. The same mechanism can be used to support information resource discovery. Context comparison leads to changes in schema correspondences that are used to support information correlation.

In a multi-database system, schematic conflicts between two objects are usually of interest only when the objects have some semantic similarity. In this paper we try to reconcile the schematic and semantic perspectives. We introduce a uniform formalism called schema correspondences to represent structural similarities between the objects. We represent the semantic similarities between the objects using the concept of semantic proximity. We show how the reconciliation is achieved by illustrating the association of the schema correspondence(s) with and as component(s) of the semantic proximity. We also provide a data model independent semantic taxonomy on the basis of the semantic proximity defined. We then enumerate and classify the schematic and data conflicts. The association between the schema correspondences and semantic proximity helps represent the possible semantic similarities between two objects having these conflicts. One representation of uncertain information using semantic proximity as the basis is explored. Issues of inconsistent information are also discussed in the framework of semantic proximity.