Advance Technique for Early Detection of Breast Cancer Using Textual Analysis from Digital Mammogram

Abstract

The field of image processing gaining importance is not only for its rapid and continuous progress but also for accurate and advanced analysis. Mammography is the most popular imaging technique for the detection of breast cancer Anatomical structure of a lesion is obtained properly compared to other imaging modalities like CT( Computed Tomography), MRI (Magnetic Resonance Imaging), PET (Positron-emission tomography). In this work, an algorithm has been developed for the detection of breast cancer. The proposed method has consisted of three steps: preprocessing, segmentation and feature extraction. After segmentation of cancerous region, it is characterized with statistical features using first-order histogram and Gray Level Co-occurrence Matrix (GLCM)). Based on these two types of feature extraction methods, normal and cancerous mammograms have been diagnosed.