Early Detection of Breast Cancer Using Hybrid Classification Technique

In this paper, we are about to see the comparison of an existing system with the proposed system by achieving high accuracy with less time delay. Breast cancer which is a deadly disease can be seen numerous in number among women now-a-days. Mammography is considered to be one of the main tools for the early detection of breast cancer. Medical practitioners use various methods to identify the presence of micro calcifications. The medical image processing techniques are used for many medical problems and the same can be applied for the problem of cancer detection. In general, the mammogram images are used to identify the presence of cancer. But the accuracy of identifying the presence of cancer requires many constraints and cannot be easily said that there exist breast cancer. Radiologist uses the mammogram images which are not much accurate and clear. The mammogram images have only black and white pixels with various gray scale values. Even though, the presence of micro calcification has been identified, identifying the location where the exact cancer cells are present is highly a tedious job. The image processing technique for micro calcification detection has different stages namely preprocessing, segmentation and micro calcification detection. In general the mammogram images would have the presence of noise which is removed at the stage of preprocessing. In the segmentation stage, the gray scale pixels are grouped according to the value of gray scale. Based on the gray scale values the method identifies the pixels which are affected by the cancer and the shape of mass can be used for the determination to check if it is malignant or benign in the detection stage. Finally comes the classification stage and then the results.