Pixcals Statistical Based Algorithm to Detect Microcalcifications on Mammograms

Clustered microcalcifications are one of the earliest indicators of breast cancer and are detected only by mammography. For early breast cancers, screening studies suggest that 70 to 90 percent were detected based on microcalcifications alone. The present work deals with the following methodology for clustered microcalcification detection. First, the digitized film mammogram is preprocessed to reduce noise by Direct Contrast Enhancement method. Second, the subimages were extracted from full breast images. Further, Statistical Algorithm based on Gaussian Process called as PIXCALS have been developed to extract the suspicious region from the subimages.

The proposed system is able to classify an image as normal or abnormal, and also for an abnormal image it indicates the suspected area which contains microcalcifications. The images are obtained from the Digital Database for Screening Mammography (DDSM). The different kinds of images have been considered and tested using the proposed algorithm and results are encouraging.