Computer Aided Diagnosis: A Study of Bronchitis^[*)]

A Computer-Aided Learning Model (CALM) was used to recognize the characteristic symptoms of bronchitis by comparing the questionnaire responses of patients with that diagnosis to the questionnaire responses of patient without that diagnosis. The computer printed out a diagnosis profile for bronchitis which consisted of weighting values for each of 3778 questions. This diagnosis profile was used in a subsequent diagnostic trial of new patients. The computer agreed with the clinical diagnosis of bronchitis in 96°/o of cases. The study validates both the questionnaire as a means of eliciting clinically meaningful data and CALM as a basis for a computer diagnosis system.

Mit Hilfe eines computer-unterstützten Lernprogramms (CALM) zur Erfassung der charakteristischen Symptome der Bronchitis wurden Fragebogenangaben von Patienten mit der Diagnose Bronchitis mit den Angaben von Patienten ohne Bronchitis verglichen. Der Computer druckte ein Diagnosen-Profil für Bronchitis aus, welches aus gewichteten Werten für jede von insgesamt 3778 Fragen bestand. Dieses Diagnosen-Profil wurde dann in einer diagnostischen Versuchsreihe an neuen Patienten erprobt. Dabei stimmte der Computer in 96% der Fälle mit der kinisch gestellten Diagnose »Bronchitis« überein. Die vorliegende Untersuchung erweist die Brauchbarkeit des benutzten Fragebogens als Mittel zur Erfassung klinisch bedeutsamer Daten sowie des Computer-Programms CALM als Grundlage eines Computer-Diagnose-Systems.

Key-Words

Diagnosis - Learning Model - Bronchitis - Diagnosis Profile

Schlüssel-Wörter

Diagnose - Computer-Diagnostik - Bronchitis - Diagnosen-Profil

^*) Publication No. 1680 from the Walter and Eliza Hall Institute.

• References

• 1 Boyle J. A, Greig W. R, Fbanklin D. A, Harden R. McG, Buchanan W. W, McGirr E. M. Construction of a model for computer assisted diagnosis; Application to the problem of non-toxic goitre. Quart. J. Med. 35 1966; 565-588.
• 2 Brodman K, Erdmann A. J, Lorge I, Wolf H. G. Cornell Medical Index: Adjunct to medical interview. J. Amer. med. Ass. 140 1949; 530-534.
• 3 Chodosh S, Weiss E. B, Segal M. S. Chronic Bronchitis. In Conn H. F. , Conn R. B.. (Edit): Current Diagnosis. 02 1968: 151-154.
• 4 Crooks J, Murray I. P. C, Wayne E. J. Statistical methods applied to clinical diagnosis of thyrotoxicosis. Quart. J. Med. 28 1959; 211-234.
• 5 Gledhill V. X, Mathews J. D. Computer aided diagnosis: A learning model. J. Med. 01 1970; 249-264.
• 6 Gledhill V. X, Mathews J. D. Acquisition and storage of clinical data by computer, I. — The self-administered symptom history. Brit. J. Hosp. Med. (Suppl. 06) 1971; 16-18.
• 7 Gorry G. A, Barnett G. O. Sequential diagnosis by computer. J. Amer. med. Ass. 205 1968; 849-862.
• 8 Medical Research Council: Definition and Classification of chronic bronchitis. Lancet. 1965; I: 775-779.
• 9 Neurath P. W, Einstein K, Mitchell G. W. Design of a computer system to assist in differential pre-operative diagnosis for pelvic survey. New. Engl. J. Med. 280 1969; 745-749.
• 10 Overall J. E, Williams C. M. Conditional probability program for diagnosis of thyroid function. J. Amer. med. Ass. 183 1963; 307-313.
• 11 Thompson E. T, Hayden A. C. Standard Nomenclature of Diseases and Operations. New York: McGraw Hill; 1952
• 12 Van Woerkom A. J, Brodman K. Statistics for a diagnostic model. Biometrics 37 1961; 299-318.
• 13 Warner H. R, Toronto A. F, Veasy L. G. Experience with Bayes’ theorem for computer diagnosis of heart diseases. Ann. N.Y. Acad. Sei 115 1964; 558-567.