Recent figures show that approximately 1 in 11 women in the western world will develop breast cancer duringt he course of their lives. Early detection greatly improves prognosis and considerable research has been undertaken to this end. Mammographic images are di.cult to interpret even by radiologists and this makes their task error prone. One of the earliest non-palpable signs is the appearance of microcalcifications, typically 0.5 mm in diameter, representingsmall deposits of calcium salts in the breast. A novel approach to detectingm icrocalci.cations in x-ray mammography has been explored. The method is based on the use of the physics-based image representation h _int [1] and use of anisotropic diffusion to filter h _int images. The diffusion process becomes a method of detecting both noise and microcalcifications in mammograms.