Multi-valued logic in breast cancer detection
Boiculese, L.V.; Dimitriu, G.
Revista Medico-Chirurgicala a Societatii de Medici Si Naturalisti Din Iasi 107(2): 425-428
2003
ISSN/ISBN: 0048-7848 PMID: 14755954 Document Number: 560427
The aim of this paper is to determine the type of the breast cancer disease. The two classes of separation are malignant respectively benign. A multi-valued logic system (fuzzy system) was develop and applied in this classification. The system uses nine attributes as inputs that were scaled with an integer value in the range between 1 and 10. The attributes represent: 1. Clump Thickness, 2. Uniformity of Cell Size, 3. Uniformity of Cell Shape, 4. Marginal Adhesion, 5. Single Epithelial Cell Size, 6. Bare Nuclei, 7. Bland Chromatin, 8. Normal Nucleoli, 9. Mitoses. After training the system managed to get a good detection with an error less than 5%.