The paper presents the concept of using pattern recognition methodology in a medical diagnostic support system based on multi-aspect fuzzy sets theory. The essence of the proposed approach consists definition of two problems: a set of disease entity patterns and an appropriately precise mathematical formula matching the patient's medical data set (patient's health condition) to the disease entity patterns contained in the medical repository. The disease pattern and the description of the patient's condition will be presented in the form of appropriately defined multi aspect fuzzy sets. The paper presents the new mathematical modelling concept using the so-called multi-aspect fuzzy sets. The paper contains definitions of the most important characteristics of multi-aspect fuzzy sets in the context of their application in decision support algorithms. These include characteristics such as the image of the multi-aspect fuzzy set, the carrier and core, the bottom and top fronts of the fuzzy set, and many other characteristics derived from multi-criteria optimization. These concepts are illustrated with a numerical examples.