Contrast Enhancement Using CLAHE on the Wavelet Image Decomposition in Dense Breast Mammograms

Abstract

Breast cancer is a major public health problem, being the second most common type of cancer among Brazilian women. It is known that the higher the density of the breast, the greater the difficulty of assessment, which may result in a non-detection of this disease. In addition to the difficulty of evaluating dense breast images due to the similarity between normal tissue and lesion, the mammographic image presents intrinsic problems of the acquisition process, such as noise. Thus, the goal of this work is to propose a combination of techniques for filtering and contrast enhancement in dense mammographic images, evaluating them from the calculation of the signal-to-noise ratio (SNR) and the peak signal-to-noise ratio (PSNR). For denoising, it was proposed to use the wavelet transform with an automatic threshold, while for the contrast enhancement, the contrast limited adaptive histogram equalization (CLAHE) algorithm was used. The results showed that there was an increase of 50, 5% and 16, 9% for the calculation of SNR and PSNR, respectively, when the CLAHE technique was used after the denoising. As a conclusion, the use of two combined methods, one for filtering, and one for contrast enhancement, allows increasing the signal in relation to the estimated noise of the image.