Development of a New Algorithm for Detection of Mammographic Masses

Abstract

The rate of breast cancer occurrence is increasing and it is estimated that breast cancer will be the top cause of Japanese women’s cancer mortality quite soon. The examination by mammography is now becoming a major diagnostic tool for finding the cancers at an early stage. However, the number of the expert doctors is not enough in the visual interpretation of mammography and a computer-aided diagnosis (CAD) system to aid physicians is deeply required. Therefore, many research groups have developed automated schemes for detecting masses [1]–[6] and clustered microcalcifications on mammograms.