A METHOD FOR DETECTING MICROCALCIFICATIONS IN DIGITAL MAMMOGRAMS

Microcalcification clusters are often an important indicator for the d etection of malignancy in mammograms. In many cases, microcalcificatio ns are the only indication of a malignancy. However. the detection of microcalcifications can be a difficult process. They are small and can be embedded in dense tissue, This paper presents a method for automat ically detecting microcalcifications. We utilize a high-boost filter t o suppress background clutter enabling segmentation even in very dense breast tissue. We then use a threshholding and region growing techniq ue to extract candidate microcalcifications. Likely microcalcification s are then identified by a linear classifier. We apply this method to images selected from the LLNL/UCSF Digital Mammogram Library, and prod uce a receiver operating characteristic (ROC) curves to detail the tra de-off between probability of detection and false alarms. Finally, we exam the ability to properly selects threshold to achieve a desired pr obability of detection based upon a training set.