Automated detection of microcalcifications in mammograms through application of image pixel remapping and statistical filter

Mammography and finding of suspicious masses during self-examinations and clinical breast examinations form the primary screening tools for early detection of breast cancer. Mammography is essential for early detection of cancer, prior to any other means. Many abnormalities other than cancer are shown by screening mammography, leading to biopsy of the suspicious defect. Any enhancement that reduces the number of unnecessary biopsies is welcomed. Digital enhancement may aid in early detection of some patterns, such as microcalcification clusters, indicating the onset of DCIS (ductal carcinoma in situ) that accounts for 20% of all mammographically detected breast cancers, and early detection may enable a complete cure. The individual calcifications are hard to detect, due to their size and shape variability and the inhomogeneous background structure. Our study addresses only early detection of microcalcifications. We present an algorithm which locates microcalcifications based on local gray-scale variability, tissue structures and image statistics. The mammographs are digitally enhanced to accent textures and a previously developed threshold filter creates a binary mask of the calcification´s spatial location