RETRACTED ARTICLE: Dynamic cross propagation algorithm based detection of micro calcification in digital mammogram

Abstract

Micro calcification in mammograms may be considered early signs of breast cancer. However, their detection by a variety of factors is a very challenging task to find cancer in an instant starting stage. A breast compression, more difficult breast anatomy, and in some cases, inaccessible size calculations, as well as the significant variation of low contrast, inherent to mammograms. Therefore a computerized image processing scheme is implemented for detecting early-stage Microcalcification in mammograms. Necessarily the masses and microcalcification may be a prominent early symptom of breast cancer. This proposed Dynamic Cross Propagation Algorithm (DCPA) tested several images from digital database mammography for cancer research and diagnosis. In the first stage of preprocessing will proceed for reducing noise in the original image. In the second stage to Finding of interest, which can be done by using a superior region of interest (SROI) detection and region segmentation, is in the area of a suspicious mass on a mammogram. After detecting suspicious mass on the mammogram, the features like mean, standard deviation, variance, skewness, and entropy values are obtained for further classification. In the Fourth stage the feature result, based classification using the proposed Dynamic Cross Propagation Algorithm (DCPA) compiles to find the loss function in the micro image and compare it with the train the model. In this method, SROI fractal dimensions have computed the value that corresponds with training data to find the suspicious microcalcification regions. The proposed system shows a high rate of true positives and a low rate of false-positive, a superior performance compared to the conventional method. The performance of the proposed DCPA technique is analyzed in terms of precision, recall and F-measure. The experimental results produced a classification that gives better accuracy 98.38% of which indicates that the proposed system is more promising in classification digital mammograms.