Detection of lesion in mammograms
Adel Dallali, Amel Slimen, Salim El Khediri, Youssra Khemili · 2018
Breast cancer is one of the most dangerous diseases that causes countless mortals among women. In fact, the early detection of this illness is the only way to reduce the mortality rate. Several studies have been carried out, over the last two decades, to develop tools for diagnosis (CAD). In this context, we present a method to detect breast cancer based on three steps. The first one is a pre-processing stage in which histogram equalization and adaptive limited contrast are used. The second step is image segmentation relying on the application of histogram thresholding method, while the third step employing the co-occurrence matrix consists in characteristics extraction. The proposed method was tested on several mammograms from the mini-MIAS mammographic database.