Infiltrative breast cancer initial detection based on double-scale sech template matching

Breast mass is a significant indicator of breast cancer. It is difficult for CAD (Computer Aided Diagnosis, CAD) system to detect infiltrative breast masses accurately because the masses are quite subtle, often occurred in dense areas of breast tissues, change various in shape and size and always infiltrate into breast tissues. In this paper, we proposed a new algorithm based on double-scale Sech template for early infiltrative masses initial detection from mammographic images. This algorithm includes the breast region extraction, breast pectoral-muscle remove and suspected mass regions detection. The proposed algorithm was tested on a database of 60 mammograms which masses had previously been marked by experienced radiologists. The detection sensitivity in this method is 100%, and average 2.77 false positive regions in every image. These results show the effectiveness of the proposed algorithm.