Fuzzy multi-layer SVM classification of breast cancer mammogram images

A huge increase in health issues has set new challenges to clinical routine for patient's record about diagnosis, treatment and follow-up, with help of data & image processing it is possible to assist or automate the radiologist for diagnosis. Detection of breast cancer is done with mammogram, which are low dose x-ray images. Mammogram image play a totally vast function in early detection of breast most cancers. Usually photo texture analysis is used for clustering and classification primarily based on content of picture. This paper concentrated on Fuzzy-Multi layer SVM (FMSVM) classifier for evaluating the features extracted and to determine its effects. The proposed FMSVM version indicates promising consequences when compared with different classifiers used most generally within the literature and can offer a destiny for more sophisticated statistical features based most cancers prognostic models. The proposed method is evaluated on a set composed of images extracted from Mini MIAS databases. The examination over the images show that the proposed method is efficient and effective for detecting the malignant, benign and normal tumors, as well the accuracy achieved is about 98%. The outcomes show the promising factors of the proposed methodology along with the suggestions for the future work.