Rough Sets Approach to Analysis of Data of Diagnostic Peritoneal Lavage Applied for Multiple Injuries Patients

Abstract

Three kinds of information systems containing 80 patiens with multiple injuries, who had diagnostic peritoneal lavage, are analysed with the concept of rough sets. Twenty three attributes are used to describe these patients. Twenty of them concern: anamnesis, description of general condition, state of consciousness, and kind of injuries. Three last attributes define classification of patients according to three clinical problems.

Using the rough sets methodology implemented in the microcomputer program RoughDAS, the information systems are reduced so as to get the minimum subsets of attributes ensuring an acceptable quality of the classification. The relationship between these attributes and three classifications is shown by the decision algorithms Analysis of these algorithms leads to the concrete clinical conclusions.