KLASIFIKASI CITRA MAMMOGRAM MENGGUNAKAN METODE K-MEANS CLUSTERING, GLCM, DAN SUPPORT VECTOR MACHINE(SVM)

Abstract

Breast cancer is one of the non-communicable diseases that tend to continue to increase every year. The disease occurs almost entirely in women, but can also occur in men. The best way to identify the presence of breast cancer in addition to ultrasound examination can also be by interpreting a mammogram image using low doses of X-rays that can show abnormalities or abnormalities in the breast in a very small form. The cancer detection system that will be built is a detection system in the breast using mammogram imagery that will pass through the pre-processing stage, image segmentation stage, post-processing stage, feature extraction stage, and classification stage. Methods in the stages of breast cancer detection system are K-Means Clustering Method in segmentation process, GLCM in feature extraction and SVM in classification process. This system will detect and classify normal or abnormal breast cancer based on the characteristics that have been extracted, namely contrast, correlation, energy, and homogeneity by using the process of training (training) and testing (test). The accuracy achieved on this system is 85% of the 20 test images attempted.