Using fuzzy sugeno integral as an aggregation operator of ensemble of fuzzy decision trees in the recognition of HER2 breast cancer histopathology images

Abstract

In this paper a decision making support system based on fuzzy logic is considered. The examined decision problem is related to the problem of recognition of histopathology images with respect to the degree of HER2/neu receptor overexpression. We used fuzzy decision trees, defined over different sets of image features, as separate image classifiers. Then, the corresponding classifiers results were aggregated with the fuzzy Sugeno integral to make final recognition decision. The proposed approach was tested over real clinical data of HER2 breast cancer histopathology images.