Determination Of Breast Density Using Fuzzy Logic

Abstract

Background and Aim: The risk of breast cancer increases directly in line with breast density. Therefore, it is important to pay more attention to denser breasts in order to detect abnormalities. The aim of this paper was to design and suggest a quantitative method to categorize breast density in digital mammogram images using fuzzy logic. Materials and Methods: This was a crosectional study which resulted in developing a new system. The research population was including patients who undergo mammography in National Cancer Society of Malaysia during 2010 to 2011. Sample included 220 mammogram images which was selected randomly. Data analysis was done using SPSS with Kappa statistics. Results: Accuracy level of 92.8% was obtained based on evaluation of the system and there was a strong correlation between the system output and radiologists’ estimation (K=0.87, p=0.0001). Conclusion : Results obtained from the suggested system had higher performance than similar systems. Therefore, it could be concluded that the fuzzy logic may be used in this area. In addition, such systems could be helpful for physicians.

Keywords

• fuzzy logic
• breast density
• classification
• mammogram
• risk factor