Segmentation of histopathology HER2/neu images with fuzzy decision tree and Takagi-Sugeno reasoning

The Human Epidermal Growth Factor Receptor 2 (HER2/neu) is a biomarker, recognized as a valuable prognostic and predictive factor for breast cancer. In approximately 20% of primary breast cancers, the HER2/neu protein is over-expressed. By recent clinical research, a treatment procedure, with corresponding monoclonal antibodies specifically designed to target the HER2/neu receptor, was confirmed. Therefore, in modern breast cancer diagnostics, it is critical to provide accurate recognition of the HER2/neu positive breast cancer. This can be done by segmentation of the membranes of cancer cells that are visualized as HER2/neu over-expressed on images acquired from corresponding histopathology preparations. In our research, we propose an accurate segmentation process of these structures using an appropriately defined fuzzy decision tree. Moreover, we introduce a new reasoning concept based on the Takagi-Sugeno inference model.