Enhanced identification of different types of microcalcifications using the wavelet transform method

International audienceMicrocalcifications and Clustered microcalcifications are known to be the first sign of a development of an eventual cancer. They appear as small and bright region with irregular shape in the breast. Their diversity in their shape, their directionality, their size and localisation in a dense mammogram confirm the major difficulty of their detection. The aim of our scheme is the de¬velopment of a method for the detection of all type of microcalcifications. The wavelet transform has recently emerged as a powerful tool for non-stationary signal analysis, its discrete version is closely related to filter banks. Also, multiresolution signal analysis has been used in image processing. This paper uses the relationship between these two techniques to provide a method of enhancing all type of microcalcifications. The digital mammograms are first processed by the wavelet transform with its multiresolution analysis. The enhancement of the microcalcifications is assured for each different type by using for an appropriate wavelet coefficients