Abstract—The innovative approach consists of using the same algorithmic core for processing images to detect both microcalcifications and masses. Despite the advancement in the medical sciences cancer is claiming more than 50% of the people afflicted by it every year. Of all cancer incidence women around the world, the most commonly diagnosed type of non-skin cancer which results in death is Breast Cancer and this can be best detected by digital mammography. This paper includes the design and development of software expert system for real time mammogram image analysis. The system so designed would give the radiologist an idea about the exact shape and size of any tumor present in the breast. Radiologists however are unable to detect the cancerous growth when benign though it is detected in the mammograms due to varying criteria like dense flesh around the cancer or distractions due to neighboring features. This problem will be resolved by using Digital Image Processing techniques like Image Segmentation where the image will be segmented into similar regions by meaningfully assigning class labels to similar pixels in a region. Hence the cancerous growth will be detected in its early stage and the Radiologists will be able to do better diagnosis because Image Segmentation techniques are simple yet very effective. In this paper an innovative method is applied which consist mainly three steps. In the first step normalizes the regions in the breast images through uniform distribution of histogram equalization. In the second step fuzzy logic is applied to remove ambiguity in the misclassification region and in the third step a new Weight is applied to the previously extended OTSU method.
Index Terms—Breast Cancer, Mammogram, ROI, OTSU Threshold, Histogram Weight, Mean, Variance
Cite: C. Naga Raju, C. Harikiran, and T. Siva Priya, "Design of Primary Screening Tool for Early Detection of Breast Cancer," Journal of Advances in Information Technology, Vol. 3, No. 4, pp. 228-235, November, 2012.doi:10.4304/jait.3.4.228-235
Copyright © 2013-2020. JAIT. All Rights Reserved
This work is licensed under the Creative Commons Attribution License (CC BY-NC-ND 4.0)