Abstract
The ability to handle uncertain information in a medical expert system has been recognized since the inception of these systems over a decade ago. Early systems utilized ad hoc measures, often providing certainty factors as an indication to the user of the level of confidence in the decision. During the same period, theoretical development of techniques in approximate reasoning has proceeded rapidly. In this paper, these theoretical techniques are applied to a working expert system for the analysis of chest pain. Examples are provided which illustrate the potential benefit of this approach.