Automatic Pectoral Muscles Detection and Removal in Mammogram Images


 The main aim of the Computer-Aided Detection/Diagnosis system is to assist radiologists in examining digital mammograms. Digital mammogram most popular screening technique for early detection breast cancer. One problems analysis presence pectoral muscles region with high intensity upper right or left side Media-Lateral Oblique views images. Therefore, it important remove this muscle from image accurate diagnosis results. proposed method consists three steps. In first step, noise reduced using Median filtering. second artifacts removal and extraction are performed Otsu method. Finally, extracted removed Split Orientation Local Thresholding (SOLTH) algorithm. For study, a total 110 images Mini-Mias database (MIAS) were used evaluate method’s performance. experimental results automatic observed by radiologist showed 90.9% accuracy acceptable results.