Paper: Symbolic decision support in medical care

  • Authors:
  • J. Huang;J. Fox;C. Gordon;A. Jackson-Smale

  • Affiliations:
  • Advanced Computation Laboratory, Imperial Cancer Research Fund, 61 Lincoln's Inn Fields, London WC2A 3PX, UK;Advanced Computation Laboratory, Imperial Cancer Research Fund, 61 Lincoln's Inn Fields, London WC2A 3PX, UK;Advanced Computation Laboratory, Imperial Cancer Research Fund, 61 Lincoln's Inn Fields, London WC2A 3PX, UK;Advanced Computation Laboratory, Imperial Cancer Research Fund, 61 Lincoln's Inn Fields, London WC2A 3PX, UK

  • Venue:
  • Artificial Intelligence in Medicine
  • Year:
  • 1993

Quantified Score

Hi-index 0.00

Visualization

Abstract

Symbolic decision procedures offer a flexible alternative to classical quantitative procedures for decision making, particularly when precise parameters (such as probabilities) are hard to estimate. One such procedure, based on a logic of argumentation, is described. Specifications of inference methods for such functions as proposing and refining decision options, deducing and inheriting arguments for and against options, and selecting among altematives are presented. These exploit declarative models for patient data, domain and task knowledge. A simple method for translating the specifications into executable Prolog is described. A practical and efficient toolset for using the procedure in a wide range of clinical environments is being developed within the DILEMMA project of the European Commission's Advanced Informatics in Medicine research programme.