Breast Cancer Detection using Intuitionistic Fuzzy Histogram Hyperbolization and Possibilitic Fuzzy c-mean Clustering algorithms with texture feature based Classification on Mammography Images

TLDR: This paper designates the digital execution of a model, based on an intuitionistic fuzzy histogram hyperbolization and possibilitic fuzzy c-mean clustering algorithm for early breast cancer detection, which was applied in classifier to detect about the presence of cancerous tumor in mammogram images.

Abstract: During past 20 years, it is stated that cancer belongings are mounting all-inclusive. Amid innumerable natures of cancers, breast cancer is witnessed as key reason of demise among women. Ultrasound, x-ray (mammograms and x-ray computed tomography), magnetic resonance imaging, thermography and nuclear medicine functional imaging are different modalities offered for early stage breast cancer detection. Mammography technology is a unadventurous breast cancer practice that can perceive tumorous masses on lower cost and better truthfulness. This paper designates the digital execution of a model, based on an intuitionistic fuzzy histogram hyperbolization and possibilitic fuzzy c-mean clustering algorithm for early breast cancer detection. Clustering plays a key role in segmentation fragment. Classical fuzzy clustering assigns data to multiple clusters at different degrees of membership but irrelevant data are also allocated to some clusters that do not relate to them. In our newfangled work we bound possibilistic method with fuzzy c-mean to resolve this issue after applying intuitionistic fuzzy histogram hyperbolization algorithm in initial preprocessing phase in the mammogram images. Further texture feature extraction technique is used for extracting features. Developed rules was applied in classifier to detect about the presence of cancerous tumor in mammogram images. The inclusive classification accuracy achieved 94% during training stage.