Towards the Unification of Inference Structures in Medical Diagnostic Tasks

Abstract:

The central purpose of artificial intelligence applied to medicine is to develop models for diagnosis and therapy planning at the knowledge level, in the Newell sense, and software environments to facilitate the reduction of these models to the symbol level. The usual methodology (KADS, CommonKADS, GAMES, HELIOS, Protégé, etc.) has been to develop libraries of generic tasks and resuable problem-solving methods with explicit ontologies. The principal problem which clinicians have with these methodological developments concerns the diversity and complexity of new terms whose meaning is not sufficiently clear, precise, unambiguous and consensual for them to be accessible in the daily clinical environment. As a contribution to the solution of this problem, we develop in this article the conjecture that one inference structure is enough to describe the set of analysis tasks associated with medical diagnoses. To this end, we first propose a modification of the systematic diagnostic inference scheme to obtain an analysis generic task and then compare it with the monitoring and the heuristic classification task inference schemes using as comparison criteria the compatibility of domain roles (data structures), the similarity in the inferences, and the commonality in the set of assumptions which underlie the functionally equivalent models. The equivalences proposed are illustrated with several examples. Note that though our ongoing work aims to simplify the methodology and to increase the precision of the terms used, the proposal presented here should be viewed more in the nature of a conjecture.

Keywords

Medical Diagnosis - Inference Structures - Unification - KADS

• REFERENCES

• 1 Marr D. Vision. New York: Freeman; 1982
• 2 Newell A. The Knowledge Level. AI Magazine 1981; summer 1-20.
• 3 Maturana H. The Organization of the Living: a Theory of the Living Organization. Int J Man-Machine Studies 1975; 7: 313.
• 4 Varela FJ. Principles of Biological Autonomy. The North Holland Series in General Systems Research. North Holland: 1979
• 5 Mira J, Delgado AE. Some Comments on the Anthropocentric Viewpoint in the Neurocybernetic Methodology. Proc. of the Seventh International Congress of Cybernetics and Systems 1987; 2: 891-5.
• 6 Mira J, Delgado AE, Boticario JG, Diez FJ. Aspectos Básicos de la Inteligencia Artificial. Sanzy Torres S. L. Madrid: 1995
• 7 Puerta AR, Tu SW, Musen MA. Modelling tasks with mechanisms. International Journals of Intelligent Systems 1993; 8: 129-52.
• 8 Ford K, Bradshaw JM, Adams-Webber JR, Agnew NM. Knowledge Acquisition as a constructive modelling activity. International Journal of Intelligent Systems 1993; 8: 9-32.
• 9 Buchanan BG, Feigenbaum EA. DEN-DRAL and Meta-DENDRAL: Their Applications Dimension. Artif Intell 1978; 11: 5-24.
• 10 Clancey WJ. Heuristic Classification. Artificial Intelligence 1985; 27: 289-350.
• 11 Clancey WJ. Notes on: Heuristic Classification. Artificial Intelligence 1993; 59: 191-6.
• 12 Chandrasekaran B. Towards a Taxonomy of Problem Solving Types AI Magazine. 1983; 4 (1) 9-17.
• 13 Chandrasekaran B. Generic Tasks in Knowledge Based Reasoning: High-Level Building Blocks for Expert System Design. IEEE Expert 1986; 1: 23-9.
• 14 Chandrasekaran B, Johnson TR, Smith JW. Task-structure analysis for knowledge modeling. Comm ACM 1992; 35: 124-37.
• 15 Fafner R. Integrated Generic Task Toolset. Internet. 1996 http://www.cis.ohio-state.edu/hypertext/LAIR/Misc/GTTooI-set/toolset.html
• 16 Breuker J, Wielinga B. Models of expertise in knowledge acquisition. In: Guida G, Tasso C. eds. Topics in Expert System Design: Methodologies and Tools. Amsterdam: North-Holland Publ., Comp; 1989: 256-95.
• 17 Wielinga BJ, Schreiber AT, Breuker JA. KADS: A Modelling Approach to Knowledge Engineering. Knowledge Acquisition (Special Issue) 1992; 4 (1) 5-54.
• 18 Tansely DSW, Hayball CC. Knowledge-Based Systems Analysis and Design. A KADS Developer's Handbook. Prentice Hall; 1993
• 19 Breuker J, Van de Velde W. eds. Common-KADS Library for Expertise Modelling. Reusable Problem Solving Components. IOS Press 1994
• 20 Van Heijst G, Falasconi S, Abu-Hanna A, Schreiber G, Stefaneiii M. A Case Study in Ontology Library Construction”. Artif Intell Med 1995; 7: 227-55.
• 21 Van Heijst G, Schreiber G, Lanzóla G, Stefanelli M. Foundations for a Methodology for Medical KBS Development. Knowledge Acquisition 1994; 6: 395-434.
• 22 Van Heijst G, Schreiber ATh, Wielinga BJ. Using explicit ontologies in KBS development. Int J of Human-Computer Studies 1997; 45: 183-292.
• 23 Fensel D, Van Harmelen F. A Comparison of Languages which Operationalize and Formalize KADS Models of Expertise. The Knowledge Engineering Review 1994; 9: 105-46.
• 24 Schreiber G, Wielinga B, Akkermans H, Van der Velde W, Anjewierden A. “CML: The CommonKADS Conceptual Modelling Language”. Internet. 1994 www.swi.psy.uva.ul/CommonKADS
• 25 Puerta AR, Tu SW, Musen MA. "Modelling Tasks with Mechanisms". International Journals of Intelligent Systems 1993; 8: 12952.
• 26 Tu SW, Eriksson H, Gennari JH, Shahar Y, Musen MA. Ontology-based Configuration of Problem-solving Methods and Generation of Knowledge-acquisition Tools: Application of PROTÉGÉ-II to Protocol-based Decision Support. Artificial Intelligence in Medicine 1995; 7: 257-89.
• 27 Steels L. The Componential Framework and its Role in Reusability. In Jean-Marc David JPK, Simmons R. eds. Second Generation Expert Systems. Berlin: Springer Verlag; 1993
• 28 GAMES. http://www.swi.psy.uva.nl/projects/GAMES-II/home.html
• 29 Cuena J, Molina M. KSM: Un entorno para la Construcción y Reutilización de Modelos de Conocimiento. In: Actas V. Congreso Iberoamericano de Inteligencia Artificial. Iberamia 96. LIMUSA Noriega eds 1996: 215-24.
• 30 Ramoni M, Stefaneiii M, Magnani L, Barosi G. An Epistemiological Framework for Medical Knowledge-Based Systems. IEEE Transactions on Systems, Man, and Cybernetics 1992; 6: 1361-75.
• 31 Falasconi S, Lanzola G, Stefanelli M. Ontology and Terminology Servers in Agent-based Health-care Information Systems. Methods of Information in Medicine 1977; 36: 30-43.
• 32 Benjamins VR, Aben M. Structure-preserving knowledge-based system development through reusable libraries: a case study in diagnosis. Int J Human-Computer Studies 1997; 47: 259-88.
• 33 Barreiro A, Otero RP, Marin R, Mira J. A Modular Knowledge Base for the Follow-Up of Clinical Protocols. Methods of Information in Medicine 1993; 5: 378-81.
• 34 Nathan DG, Oski FA. Hematology of Infancy Childhood, 3rd. ed. Philadelphia: WB Saunders; 1987
• 35 Porter BW, Bareiss R, Holte RC. Concept Learning and Heuristic Classification in Weak-Theory Domains. Artificial Intelligence 1990; 45: 229-63.
• 36 Barreiro A, Otero RP, Marin R, Delgado AE, Mira J. Follow-up of patients included in oncology protocols using Medtool. In: Proceedings V Mediterranean Conference on Medical and Biological Engineering. MEDI-CON 89. Patras 1989; 280-1.
• 37 White R, Grant A, Cramp DG. Small cell lung cancer protocol comparison. TAO Esprit Project 1592 Internal Report. March 1989
• 38 Mira J, Otero RP, Barro S, Barreiro A, Ruiz R, Delgado AE, Marin R, Amaro A, Jacket M. On knowledge-based system in ccu's: monitoring and patients follow-up. In: Knowledge Based Systems in Medicine: Methods, Applications and Evaluation. Lecture Notes in Medical Informatics 47. Talmon JL, Fox J. (eds) Berlin: Springer-Verlag; 1991: 174-92.