A Novel Breast Image Preprocessing For Full Field Digital Mammographic Segmentation and Risk Classification

To obtain optimal breast image quality during the image acquisition, a compression paddle is used to even the breast thickness. Clinical observation has indicated that breast peripheral areas may not be fully compressed, and may cause unexpected intensity and texture variation within these areas. Such breast parenchymal appearance discrepancies may not be desirable for tissue modelling within computer aided mammography. This paper describes a novel mammographic image preprocessing method to improve the image quality before analysis. Mammographic segmentation and risk classification were performed to facilitate a quantitative and qualitative evaluation, using digital mammographic images. Visual assessment indicated significant improvement on segmented anatomical structures and tissue specific areas when using the processed images. The achieved risk classification accuracies are 80% and 79% for Birads and Tabár risk scheme, respectively. The developed method has demonstrated an ability to improve the quality of mammographic segmentation, leading to more accurate risk classification. This in turn can be found useful in early breast cancer detection, risk-stratified screening, and aiding radiologists in the process of decision making prior to surgery and/or treatment.