The practice of medicine basically depends on three general concepts: an individual patient image, the individual disease course of this patient, and an expectation of the future development of the disease. Physicians use heuristics for the cognitive processing of these concepts. Basic clinical heuristics are put into three algorithms: a history-based-learning algorithm, a diagnostic-inferencing algorithm, and a prognostic-planning algorithm. A proto-theory of clinical reasoning for practicing physicians is proposed.
Ebooka przeczytasz w aplikacjach Legimi na:
Liczba stron: 90
Odsłuch ebooka (TTS) dostepny w abonamencie „ebooki+audiobooki bez limitu” w aplikacjach Legimi na:
To Peter Nissen
EXERCISES: TOOLS FOR COGNITIVE PROCESSING
The text is the core of a knowledge-engeneering lecture in computer science (AI-working group, Technical Informatics, Universitiy Bielefeld, 1992-95) introducing master students to basic reasoning principles in clinical medicine. The very general approach to the practice of medicine oversimplifies the complexity of real clinical settings. Many of the issues would require extended discussions. The text provides a proto-theory.
The author is much in dept to Reinhard Flessner for reading of the text and his criticisms.
In our professional medical language the term “clinical medicine” denotes all the activities of solving individual health problems under real world conditions. Clinical medicine has often been analysed under the focus of processing a nosological taxonomy for the classification of individual disease cases. From the viewpoint of medical practice cognitive processing is more than solving classification tasks.1 In contrast the core structure of clinical reasoning is the dynamic concept of the individual disease course. Physicians use basic clinical heuristics to reconstruct idiosyncratic disease courses. Both elements, disease course and reasoning heuristics, are necessary to apply medical knowledge to the real world problems of the patients.
My thesis is that the practice of medicine basically depends on three general concepts: an individual patient image, the individual disease course of this patient, and an expectation of the future development of the disease. I analyze the logical structures of these concepts in chapter 3 (patient image) and chapter 4 (disease course).
Physicians use heuristics for the cognitive processing of these concepts under real world conditions. Heuristics have the advantage that they guide medical decisions within fuzzy boundaries. Heuristics are open for new information, for intuitions.2 In chapter 5 basic clinical heuristics are condensed into three algorithms: a history-based-learning algorithm, a diagnostic-inferencing algorithm, and a prognostic-planning algorithm. Chapter 6 summerizes the results of my analysis.
Although computer-based medical information processing has tremendously progressed in the last years, there is no convincing penetration of these technologies into our everyday practice of medicine. Chapter 7 enumerates possible applications of the proposed theory. A list of reasonable exercises may encourage cognitive engineers to design tools to support cognitive processing of individual patient data. These tools should realize the proposed algorithms. They should be judgment assistants or “tools for thinking”3 for the practicing physician.
1 Braunwald E, Isselbacher KJ, Petersdorf RG, Wilson JD, Martin JB, Fauci AS (eds.). Harrison´s Principles of Internal Medicine, New York, et al.: McGraw-Hill Book Company, 11th ed. 1987. Hurst, J.W. (ed.), Medicine for the Practicing Physician, 2nd edition, Boston et. al.: Butterworth, 1988. Kahneman D, Slovic P, Tversky A (eds.). Judgement under Uncertainty: Heuristics and Biases. Cambridge, England: Cambridge University Press 1982. Macleod John (ed.). Clinical Examination. Edingburgh, London, NewYork: Churchll Livingstone, 1979. Murphy EA. Probability in Medicine. Baltimore, London: The Johns Hopkins University Press, 1979. Murphy EA. Skepsis, Dogma and Belief. Uses and Abuses in Medicine. Baltimore, London: The Johns Hopkins University Press, 1981.
2 Gigerenzer G. Gut Feelings. New York: Viking. 2007. Gigerenzer G. Fast and Frugal heuristics: tools of bounded rationality, in: Koehler D, Harvey N (eds.): Blackwell handbook of judgement and decision making. Oxford 2004, 62-88. Schwartz S, Griffin T. Medical Thinking. The Psychology of Medical Judgment and Decision Making, Springer: New York, 1986.
3 Hurst, J.W. (ed.), a.a.O., p13.
The author was interviewed as a medical expert during the knowledge acquisition process for the HYPERCON-project (HYPERCON 1992 – 1995: University of Bielefeld, Faculty of Technology, Knowledge Based Systems - Artificial Intelligence Research Group, Head: Ipke Wachsmuth).4 The reconstruction of general structures in clinical medicine is based upon the transcripts of these interviews.
The formal approach for the reconstruction uses Patrick Suppes´s method of axiomatizing theories by defining set theoretical predicates.5 His informal set theoretic approach has later been extended to the so called “Structuralist View of Scientific Theories”.6 Following these ideas the basic structures in clinical medicine are reconstructed. The technique of defining set theoretical predicates.is self-explanatory, the formal representation uses naïve set theory.
This approach has three advantages: (1) The predicates reconstruct and summarize basic structures and their functions in clinical medicine. (2) They can be used to elicit basic heuristics of medical knowledge processing. (3) During the initial phase of a knowledge engineering process this kind of intuitive representation helps to axiomatize the medical domain in a precise manner without recourse to specialized formal languages which later on are necessary to transform the modelling into computer programs.
4 Heller B, Meyer-Fujara J, Schlegelmilch S, Wachsmuth I. HYPERCON: ein Konsultationssystem zur Hypertonie auf der Basis modular organisierter Wissensbestände. University of Bielefeld ; Fakultäten. Technische Fakultät, in: G. Barth et al.: Anwendungen der künstlichen Intelligenz: KI-94, 18. Fachtagung für Künstliche Intelligenz, Saarbrücken, 22./23. September 1994, Berlin: Springer: 1994, 155-169. Heller B. Modularisierung und Fokussierung erweiterbarer komplexer Wissensbasen auf der Basis von Kompetenzeinheiten. Bielefeld, Univ., Diss., 1995. Sankt Augustin: Reihe: Dissertationen zur künstlichen Intelligenz. 1996. Heller B, Herre H, Lippoldt K, Loeffler M. Standardized Terminology for Clinical Trial Protocols Based on Top-Level Ontological Categories. In: Kaiser K et al (eds.) Computer-based Support for Clinical Guidelines and Protocols, IOS Press 2004, 46-60. Meyer-Fujara J, Heller B, Schlegelmilch S, Wachsmuth I. Knowledge-level modularization of a complex knowledge base, in: Bernhard Nebel et al. Advances in artificial intelligence: proceedings, eds. (Lecture notes in computer science; 861), Berlin: Springer, 1994, 214-225. Müller U. SCREEKON: Medizinisches Konsultations-system für Screeningfunktionen am Beispiel des therapeutischen Managements von Patienten mit cerebralen Durchblutungsstörungen, Angewandte Inforamtik 2/1989, 76. Müller-Kolck U. Medizinische Therapieentscheidungen mit SCREECON, in: Savory SE (ed.) Expertensysteme: Nutzen für ihr Unternehmen. Ein Leitfaden für Entscheidungsträger. München, Wien: R. Oldenbourg; 1989. Müller-Kolck U. Expert system support for the therapeutical management of cerebrovascular disease. Artificial Intelligence in Medicine, 2: 1990, 35-42. Müller-Kolck U. Diagnostic Consultations: A Speech-Act Theoretical Reconstruction fort he Design of Consultations Systems. Methods of Information, 1991, 311-315. Müller-Kolck U. Expertensysteme als metadiagnostische Hilfsmittel in ärztlichen Entscheidungsprozessen, in: Hucklenbroich P, Toellner R (eds.) Künstliche Intelligenz in der Medizin. Klinisch-methodologische Aspekte medizinischer Expertensysteme, Stuttgart, Jena, New York: Fischer Verlag, 1993, 141-159. Müller-Kolck U. Basic structures of nosology in medical arguments (in german), in: Meggle G., Nida-Rümelin J., Perspektiven der Analytischen Philosophie, Band 18, Proceedings of the 2nd Conference „Perspectives in Analytical Philosophy“, Vol III, Berlin, New York: De Gruyter 1997, 518-528. Müller-Kolck U. Modellierung individueller Prognosen in der klinischen Medizin. Theory Biosci, 120: 2001, 45-56.
5 Suppes P., Introduction to Logic. New York, Cincinnati, Toronto,London, Melbourne: Van Nostrand Reinhold Company, 1957.
6 Balzer, W., Moulines C. U., Sneed, J.D.: An Architecture for Science, The Structuralistic Program, Dordrecht: D. Reidel Publ., 1987.
This chapter introduces the basic structures for reasoning in the practice of clinical medicine. To simplify matters some abstract terms like “patient self image”, “partial potential patient image” or “potential patient image” are proposed. They are explicated on a very general level. The “image”-term stresses the “iconographic” view of the physician on his patient. From the viewpoint of cognitive science these patient images may be interpreted as the physician´s internal cognitive representations of individual patient data.
The iconographic view is predominant in clinical reasoning. The interaction between patient and physician is based on images of the patient. The physician produces multiple images to capture the patient’s problem. The old and often used analogy between physicians and artists illustrates the intuitive meaning of the iconographic view. This analogy compares the view of the physician on his patient with the view of the painter on his model. The painter creates an individual image. He transforms his sensual cognitive impressions into forms and colours on the screen, the portrait.
The portrait of the model resembles the medical patient image whose features are documented by the physician in multiple clinical data records. Interactions between physicians and their patients are based on similar iconographic acts. Like the painter the physician portraits his model, i. e. his patient. The physician creates an internal cognitive image representation. These images are often classified as mosaics, because they are puzzled together from single pieces of medical data. The physician uses incomplete and fuzzy clinical data sets to create a picture of the ill subject in its individual domain. This approach is called the axiom of iconographic reasoning in clinical medicine.
All interactions between a physician and his patient are influenced by their self-images. First, the patient has subjective intentions. He wants to achieve a specific result by consulting the doctor. The physician on the other side produces his personal individual image of the patient. This image depends on the doctor’s own intentions, preferences, on his medical knowledge, and on his technical skills. In clinical medicine all interactions are deeply rooted in both the patient’s self-image and in the doctor’s self-image. This is the axiom of interacting self-images.
The analysis of the basic structures of clinical reasoning starts with the patient’s self-image, abbreviated with PSI, and the doctor’s self-image, abbreviated with DSI.
Patients are often vague in their communications. Issues and facts of their problems are hidden by his / her personal interpretations of disease signs. Fears and worries disguise his true situation. Successful communication depends on the linguistic abilities of the patient. On the one hand, extremes of such communications are illustrated by a person, who exaggerates his problem in a never ending tale of woe and, on the other hand, by a disabled person, who is unable to speak or to act. In the first case, the physician has to listen to the patient and has to separate relevant from less relevant facts. In the second case he observes the patient for a longer time and consults medical attendants and nurses to get a correct image of the patient’s problem. Thus, the physician is confronted with open and hidden contexts. He has to understand the individual life situation of the patient.
History taking is the basis of clinical medicine. In contrast to scientific investigations, which refer to the description of classes, historical reflections refer to individual events or individual cases. History taking leads to a first image
Tysiące ebooków i audiobooków
Ich liczba ciągle rośnie, a Ty masz gwarancję niezmiennej ceny.
Napisali o nas:
Nowy sposób na e-księgarnię
Czytelnicy nie wierzą
Legimi idzie na całość
Projekt Legimi wielkim wydarzeniem
Spotify for ebooks