KLASIFIKASI CITRA MAMMOGRAM BERDASARKAN EKSTRAKSI FITUR TEXTUR DENGAN MENGGUNAKAN METODE K-NEAREST NEIGHBOUR

Abstract

Abstract – Breast cancer is a kind of cancer wich is majority by woman. Mammography has become a general method which is used to early detection to the breast cancer. In the medical field, image classification using k-NN method is used to aid radiologists in the retrieval of images with similar contents. This  methods are usually developed for specific features of images, so that those methods are not readily applicable across different kinds of medical images. This study proposes a sound methodology for CBIR of mammograms. The information can be obtained from the mammogram to be descriptors of the system. The shape and texture is some information that could be used. Result of shape feature extraction on mammograms can be seen visually, because of this feature extraction based on shape mammogram itself. While to obtain texture features from mammograms, required the formation of matrix gray level co-occurrence. From that matrix, 14 haralick texture features is got. Because in the image searching  process, first performed the normalization process in the calculation of distances for each feature. The range value of the normalization is 0-1. Weight determination of the search results for each feature is calculated by normalizing the results. This system will display the search results based on image sequence with the smallest distance value of query images. The results show that the system is able to perform retrieve image content based on mammograms for 48% precision. To process the expected future further research to improve image classification results.

Keywords: Classfication k-NN, Mammogram, Haralick, Texture features, GLCM.