Skip to Main Content
This paper presents an approach to detect small breast cancer features in mammograms such as microcalcifications, employing wavelet based methods. Microcalcifications can be difficult to locate, especially in low contrast mammographic images, such as those from younger women which have denser breast tissues. The aim is to improve contrast in mammographic images, in order to facilitate its interpretation by radiologists. The methods used involve image de-noising, wavelet image analysis and image enhancement by local adaptive operators integrated in the wavelet domain. The image is decomposed in sub-bands, the low-frequency sub-band is suppressed and than we reconstruct the image from the high-frequency sub-bands. Preliminary results indicate that the performance of this approach is acceptable. However, further studies are needed, namely to investigate the possibility of automatic or semi-automatic detection and classification of microcalcifications and image processing for regions of interest.