A framework for distributed mediation of temporal-abstraction queries to clinical databases
Section snippets
Introduction: the need for integration of clinical data and knowledge
Most clinical tasks require measurement and capture of numerous patient data of multiple types, often on electronic media. Making diagnostic or therapeutic decisions requires context-sensitive interpretation of these data. Most stored data include a time stamp in which the particular datum was valid. Temporal trends and patterns in clinical data add significant insights to static analysis. Thus, it is desirable to automatically create abstractions (short, informative, and context-sensitive
IDAN: a modular distributed temporal-abstraction mediator—an overview
IDAN is a conceptual and practical architecture that fully implements the temporal-abstraction mediation approach. In this section, we will give an overview of IDAN's components and their inter-relations. In the next section, we will describe in detail each of the IDAN components. In order to clarify the following discussion, we will start by defining several useful terms.
Temporal abstraction distinguishes among three types of objects: concepts, subjects and time intervals (Fig. 1). A concept
A detailed description of the IDAN architecture's components
In the following sections, we describe in details the methods and components used in the IDAN architecture. In Section 5, we present the main advantages (and potential disadvantages) of the architecture, and the future enhancements planned for the architecture.
Evaluation of the IDAN framework
A preliminary evaluation of the functional aspects of the IDAN architecture was performed by querying a clinical database through the interface of the KNAVE-II interactive visual exploration module [26]. The evaluation included also the usability aspects of the KNAVE-II interface, and was carried out by our collaborators in the Palo Alto, CA, USA, Veterans Administration Health Care Center, and is described in more detail elsewhere [28].
In the IDAN/KNAVE-II preliminary evaluation, the IDAN
Discussion
Several previous approaches, which have answered the need for integration of clinical data from various sources, such as that of Xu et al. [2], have focused on fusion of the data from multiple heterogeneous sources, but referred the actual abstraction to the requesting application. However, such a mechanism, although quite useful for data access, still puts the specialized task of temporal abstraction squarely on each application's shoulders. In particular, it is neither specific to the
Acknowledgments
This research was supported in part by NIH award No. LM-06806 and Israeli Ministry of Defense Award No. 89357628-01. We want to thank Dr. Mira Balaban for her help in the formal definition of the TAR language; the staff of the medical-informatics laboratory in the Ben-Gurion University for their useful comments, Samson Tu and Martin O’connor for useful discussions regarding the Chronus-II and RASTA systems, and Drs. Mary Goldstein, Susana Martins, Lawrence Basso, Herbert Kaizer, Aneel Advani,
References (35)
A Framework for knowledge-based temporal abstraction
Artif Intell
(1997)- et al.
Knowledge-based temporal abstraction in clinical domains
Artif Intell Med
(1996) - Barja ML, Bratvold T, Myllymaki J, Sonnenberger G. Informia: a mediator for integrated access to heterogeneous...
- et al.
Component-based mediation services for the integration of medical applications
Artif Intell Med
(2003) - et al.
Integration of temporal reasoning and temporal-data maintenance into a reusable database mediator to answer abstract, time-oriented queries: The Tzolkin System
J Intell Inform Syst
(1999) Mediators in the architecture of future information systems
IEEE Comput Mag
(1992)- et al.
The conceptual basis of mediation services
IEEE Expert
(1997) Patterns of enterprise application architecture
(2002)- et al.
Applying temporal abstraction in medical information systems
Ann Math Comput Teleinform
(2003) - et al.
Temporal deductive databases
Advancing the state of data integration in healthcare
J Healthcare Inform Manage
Cited by (49)
Increasing and standardizing quality of care using computerized guidelines for clinical decision support
2022, Personalized Health Systems for Cardiovascular DiseaseA method for the graphical modeling of relative temporal constraints
2019, Journal of Biomedical InformaticsAssessment of a personalized and distributed patient guidance system
2017, International Journal of Medical InformaticsAn architecture for a continuous, user-driven, and data-driven application of clinical guidelines and its evaluation
2016, Journal of Biomedical InformaticsCitation Excerpt :As to the implementation choice for specific components: Since we had rigorously evaluated our Digital Guideline Electronic Library (DeGeL) architecture [6–9,38], we used it to retrieve procedural GLs and specific declarative knowledge items when applying GL with the PICARD DSS engine. Similarly, we used a variation of the IDAN temporal-abstraction mediator [39] to answer guideline-specific temporal-abstraction knowledge-based queries to the time-oriented clinical data. Finally, to provide access to an existing EMR we used MEIDA medical-data access system [40].
Thirty years of artificial intelligence in medicine (AIME) conferences: A review of research themes
2015, Artificial Intelligence in MedicineA multiple-scenario assessment of the effect of a continuous-care, guideline-based decision support system on clinicians' compliance to clinical guidelines
2015, International Journal of Medical InformaticsCitation Excerpt :The Picard DSS engine uses our Digital Guideline Electronic Library (DeGeL) architecture [22] to retrieve procedural GLs and specific knowledge items when applying a GL. Our IDAN temporal-abstraction mediator [23] is used to answer GL-specific queries regarding knowledge-based interpretations of the time-oriented clinical data. Currently, the Picard framework serves as the backbone of the MobiGuide European Union project [24], and is used to send alerts and personalized GL-based recommendations to chronic patients (based on data from bio sensors on the patients or at their home), through the patients’ mobile devices, or to support the decisions of the patients’ care providers at the ambulatory clinic or hospital.