AUTOMATIC DETECTION OF BREAST BORDER AND NIPPLE IN DIGITAL MAMMOGRAMS

Advances in the area of computerized image analysis applied to mammogr aphy may have very important practical applications in automatically d etecting asymmetries (masses, architectural distortions, etc.) between the two breasts. We have developed a fully automatic technique to det ect the breast border and the nipple, this being a necessary prerequis ite for further image analysis. To detect the breast border, an algori thm that computes the gradient of gray levels was applied. To detect t he nipple, three algorithms were compared (maximum height of the breas t border, maximum gradient, and maximum second derivative of the gray levels across the median-top section of the breast). A combined method was also designed. The algorithms were tested on 156 digitized mammog rams. The breast segmentation results were evaluated by two expert rad iologists and one physicist. In 89% of the mammograms, the computed bo rder was in close agreement with the radiologist's estimated border. S egmentation results were acceptable to be used in computer-aided diagn ostic schemes. The mean distance between the position of the nipple in dicated by two radiologists by consensus and the position calculated b y the computer was 6 mm.