Temporal representation of clinical algorithms using expert-system and database tools. Academic Article uri icon

Overview

abstract

  • The HyperLipid Advisory System combines a rule-based implementation of a clinical algorithm (the NIH Cholesterol Education Program Expert Panel recommendations) with a temporal representation that facilitates reasoning over time while maintaining efficient storage in a standard database. The temporal representation consists of objects that model point events such as visits and interval events such as specific therapies. These objects are combined into abstractions called phases, which correspond to higher level clinical concepts such as a diet or drug treatment. The time-oriented data objects are referenced in the rules using a custom-tailored operator query language. Between user sessions relevant clinical data are stored in external files. When the advisory system is reconsulted, this information is retrieved and mapped back into an object-oriented format. Use of a commercially available expert-system shell for such tasks allows algorithm implementation in standard personal computing environments.

publication date

  • June 1, 1990

Research

keywords

  • Algorithms
  • Decision Making, Computer-Assisted
  • Expert Systems
  • Information Systems

Identity

Scopus Document Identifier

  • 0025301773

Digital Object Identifier (DOI)

  • 10.1016/0010-4809(90)90018-8

PubMed ID

  • 2350959

Additional Document Info

volume

  • 23

issue

  • 3