A Fourier-Based Algorithm for Micro-Calcification Enhancement in Mammographic Images

Breast cancer is the most widespread cancer in women in the world; it manifests mostly in two forms: microcalcifications and massive lesions. These two forms differ in density, size, shape and number. Consequently, there are two different kinds of mammographic CAD algorithms: those for microcalcifications detection, and those for massive lesions detection. The microcalcifications detection is a hard task, since they are quite small and often poorly contrasted against the background, especially in images affected by digitization noise. In a CAD system the ROI Hunter plays an important role, because missed microcalcifications at this level are definitely lost. For this reason, highlighting methods for suspected microcalcifications may be useful in a CAD system. In this work, we describe a Fourier Transform based microcalcifications enhancement method, which takes place after the image preprocessing and before the ROI Hunter step, aimed at highlighting suspected areas