These tools will no longer be maintained as of December 31, 2024. Archived website can be found here. PubMed4Hh GitHub repository can be found here. Contact NLM Customer Service if you have questions.


Pubmed for Handhelds

PUBMED FOR HANDHELDS

Search MEDLINE/PubMed

  • Title: "Auctoritas" psychiatric expert system shell.
    Author: Kovács M, Juranovics J.
    Journal: Medinfo; 1995; 8 Pt 2():997. PubMed ID: 8591608.
    Abstract:
    We present a short description of a complex psychiatric computer expert system, including functions that help the physicians and the hospital staff in the administrative, diagnostic, therapeutic, statistical, and scientific work. There are separate data-storing, health insurance-supporting, or simple advisory programs, but we can not avail a system--in our country--that provides us with all these functions together. Hence the aim of our program is to produce a universal computer system that makes the patients' long distance follow-up possible. Our diagnostic expert system shell, which is appropriate for using the symptoms and criteria scheme of the internationally accepted diagnostic systems such as DSM and ICD, helps to archive homogeneous, up-to-date psychiatric nosology; this is essential for the correct diagnostic, statistical, and scientific work. Let us introduce our expert system. It consists of four parts: administration, diagnostic decision support system, activities concerning treatment, and statistics. The part called "Administration" contains all data about actual and emitted in-patients and out-patients, including their particulars and data necessary for health insurance (duration of treatment, diagnosis); here we find and edit medical documents. The most important part of the "Auctoritas" system is the "Diagnostic decision support system." In practice, expert systems use decision trees with yes-no logic, fuzzy logic, and pattern matching on the basis of the method of deduction; and backward chaining or forward chaining on the basis of the direction of deduction. Our system uses the methods of fuzzy logic and backward chaining. In other medical disciplines, good results are achieved by applying the pattern matching method; to make validity and verification researches, however, these systems are inappropriate. The diagnoses relying on the up-to-date psychiatric diagnostic systems--DSM-IV and ICD-X--are based on classical logic and can be correctly validated and verified. Hence we have chosen the fuzzy logic, which is the up-to-date extension of classical logic and influences the validity and verification researches, for the construction of our system. The diagnostic part is a shell that can be filled up optionally with knowledge bases of the DSM-IV, ICD-X, or other diagnostic systems and has the following structure. The diagnostic course is biphased as we can differ symptoms and criteria (duration of the illness, aethyological factors). We managed to extend the traditional applications using yes-no logic with three factors that make the system more sensitive and flexible. These are the scaling, sorting by importance of symptoms, and reliability-validity results. The "scaling" means that the physician scales the input symptoms by severity; this influences the statistical probability of the possible diagnoses. "Sorting by importance" is gauging certain symptoms by importance in a syndrome. Finally the third point, "reliability-validity results," means taking account of the latest validity values of a certain disorder of the used diagnostic system--according to the latest validity researches--and the diagnostical reliability of our expert system. The "Activity concerning treatment" is a practical part of our program that contents the examination and therapy scheduling and monitoring results. Under the point of "Statistics," we can prepare all data of the patients in various ways. In summary, the "Auctoritas" computer system is a global database managing the newly-developed advisory system; it is appropriate for managing a complete hospital network system for the continuing individual long-distance observation of patients. It collects all the necessary information of one patient in one file. The long-term benefit is that it can compile and process large amounts of information about the patients and help physicians come to scientific conclusions for research and publications.
    [Abstract] [Full Text] [Related] [New Search]