Wavelet based computer aided detection of abnormalities in digital X-ray mammograms

摘要

A method is presented for the automatic detection of microcalcification clusters in digital X-ray mammograms. This utilises the 2D wavelet transform as an efficient means of reducing unwanted image structure prior to more conventional spatial and morphological operations to find potential microcalcifications. In addition, preliminary research into a novel technique for detection of subtle low contrast masses using wavelet compression is also introduced.