J. Zvarova and M. Studeny: Information theoretical approach to constitution and reduction of medical data. International Journal of Medical Informatics 45 (1997), n. 1-2, pp. 65-74.

Abstract
In medical decision problems, one meets the problem to find the most relevant piece of information for decision making. The paper is focused on a special case of diagnostic decision making when one can measure many symptoms and signs and one has to make diagnostic conclusions. The problem is as follows. One can measure symptoms and signs at a patient and has to decide about a possible diagnosis. One knows that the symptoms and signs have different costs when they are examined. Each symptom, sign or their combination has a different predictive value for the diagnosis. The task is to find out the combination of symptoms from given data with sufficient information value for the diagnostic decision making. Moreover, one simultaneously looks for a combination of symptoms and signs with the minimal costs among those carrying sufficient information. The paper describes approaches based on information measures of statistical dependence and shows the basic idea of the program CORE (COnstitution and REduction of data) which was prepared for practical applications in medicine.

AMS classification 92C50, 94A17 (68T30)

Keywords
constitution of medical data
reduction of medical data
measure of statistical dependence
multiinformation
conditional independence

A scanned pdf copy (917kB) is available.