Estudo do algoritmo CLAHE em mamas densas

Abstract

Breast cancer is one of the main diseases affecting the female population and therefore of larger interest to the public. Mammography has a role of extreme importance since it is the main exam capable of early diagnosis, increasing the chance of the patients' survival. In order to increase the visual perception of certain elements of the image and, in that way, to obtain better results in the detection of the disease, methods of digital filtration have been created; amongst these, techniques of contrast enhancement stand out. They are divided according to where the manipulation of the contrast is applied: in the image (direct) or in the histogram of the image (indirect). The objective of this academic project is to analyse an indirect technique of contrast enhancement called CLAHE (Contrast Limited Adaptive Histogram Equalization), in heterogeneously dense mammography screenings. The analysis consists of altering predefined parameters for the technique and to apply it to the mammograms through the software MATLAB, obtaining like this the variance values and SNR (signal to noise ratio) of the processed images. In this way, through the analysis of this data and of the processed mammographies, it is possible to conclude that the contrast limit enhancement of 0,01, size of tiles of [12 12] and exponential distribution are the parameters that most contributed to contrast enhancement and thus to a better understanding of the operation of CLAHE in the images.