Knowledge-Based Systems. Methods for Developing and Evaluating Expert Systems: Closing the Loop on Diagnostic Decision Support Systems

摘要

Diagnostic decision support systems typically allow a user to enter a set of manifestations and then through the use of a database, and rules or a set of algorithms produce a list of possible diagnoses which may explain one or more of the manifestations entered. This list of diagnoses is helpful as a reminder to the clinician of entities which are rarely seen in their practice, or of uncommon presentations of common disorders. The diagnoses are pointers to where a clinician might look for information on how to work up a patient. Since the clinician was reminded of a diagnosis which was not already considered, he/she may wish to look up the disorder in a trusted source. We have developed a semi-automated approach to providing electronic links between expert systems and textbooks. By providing the ability to query the text by a clinical case, the expert system provides a value added access path to within the information contained in a textbook. By providing database directed queries, expert systems can be used to automatically limit the retrieval set from a query, based on the information about disorders contained within these knowledge bases. These access paths are unique to diagnostic decision support systems and takes advantage of their database and internal logic to provide more specific access into electronic textbooks. By supplying these links we have in effect closed the loop between the patient, the information acquisition activities, the organization of a differential diagnosis, and the patient work-up.