Digital mammogram enhancement based on automatic histogram clipping

Several studies confirmed the severity of breast cancer as most mortal in women, worldwide. Premature discovery and diagnosis of cancer of breast is of significance importance in the treatment option and increased patients' possible survival opportunity. Image enhancement is one of the frequently applied techniques to curtail lethal rate by providing enhanced image, which would aid early detection and diagnosis of cancer tumor. Image enhancement is applied on the mammogram images to reduce the speckle noise and increase the contrast of the image. In this research work, it is proposed to use a novel Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm which estimates the Clip Limit adaptively by using Otsu's Method in order to enhance mammogram images. Two different threshold calculations are proposed and the proposed methods are compared with a Fuzzy Logic based adaptive clip limit CLAHE method. The experimental images were obtained from mini-MIAS mammogram database. Experiments were carried out for three different breast types; namely fatty, fatty glandular and dense glandular. The subjective test results indicate that to detect breast cancer at its earliest stage, there is need during analysis and diagnosis of the breast cancer to use both of the images obtained with the two proposed methods.