Genetic Algorithm optimized SVM for Tumor Prediction in Mammogram
1. Research Scholar, Department of CSE, Hindusthan College of Engineering and Technology, Coimbatore,
Tamilnadu, India-641 032. firstname.lastname@example.org
2. Professor & Head, Department of Computer Science and Engineering,
Hindusthan College of Engineering and Technology, Coimbatore, Tamilnadu, India-641 032.
Abstract: Screening mammography is the best available radiological technique for early breast cancer detection. But due to the huge number of mammograms requiring analysis, radiologists can make false detections. Hence, new solutions regarding automatic detection pertaining to analysis problems should be explored. Microcalcification detection/segmentation helps computerized mammogram screening to classify clusters as either malign or benign. In this paper, Gabor filter with Walsh Hadamard Transform (WHT) is applied to extract microcalcification features from mammograms. This was tested through the use of Mammographic Image Analysis Society (MIAS) mammographic databases. The mammograms were classified using a genetic-based SVM (GA-SVM) model that can automatically determine the optimal parameters, C and Gamma, of SVM with the highest predictive accuracy and generalization ability simultaneously.
[P.Ezhilarasu, Dr.J.Suganthi. Genetic Algorithm optimized SVM for Tumor Prediction in Mammogram. Life Sci J 2013;10(7s): 460-465] (ISSN:1097-8135). http://www.lifesciencesite.com. 71
Keywords: Breast cancer; Mammogram; Microcalcification; Gabor filter; Walsh Hadamard Transform; MIAS Database; Genetic Algorithm