Performance of various Denoising Techniques on Dense Glandular Mammogram

Abstract

Digital Mammogram is the major technique to detect breast cancer. Computer aided techniques helps to enhance mammogram images for better detection to radiologist. A patient who has dense breast tissue is opaque to X-ray hence mammogram is brighter such mammogram are dense glandular mammogram. When compared to fatty mammogram early finding of breast cancer is difficult in dense glandular mammogram. This paper gives a comparison study of various spatial filters used in pre-processing techniques for dense glandular, fatty glandular and fatty mammogram. Noise in the mammogram arises due to various factors such as poor illumination, capturing, movement of patient and circuit failure. Various filters are used to remove noise and improve the quality of image. Result concludes filter performance varies for different noises by comparing values of minimum mean squared error (MSE), maximum Peak signal to noise ratio (PSNR) and high structural similarities (SSI).