Enhancement of Microcalcification Features in Mammograms using Zero-Crossings of the Contourlet Transform

Department of Electronics Engineering, Govt. Model Engineering College, Thrikkakara, Kochi-682021, Kerala, India Corresponding Author: Rekha Lakshmanan ___________________________________________________________________________ Abstract Computer aided techniques can assists radiologists in detecting microcalcification, a crucial evidence in mammogram for the early diagnosis of breast cancer. A new approach is proposed in this work for early detection of breast cancer by enhancing microcalcification features using selected zero-crossings of Contourlet transform to enhance edges and to reduce image noise. Parent-child-cousin relationship among the contourlet coefficients at various levels was employed to retain the strong edge information that corresponds to the relevant features and to minimize the artifacts. Sensitivity, Specificity, contrast, Target to Background contrast ratio, Contrast Improvement Index, Average Signal to Noise Ratio and Peak Signal to Noise Ratio parameters are considered for performance evaluation of the enhancement algorithm. The mini-MIAS mammographic image database was employed for testing the accuracy of proposed method and the results were promising __________________________________________________________________________________________