KLASIFIKASI KANKER PADA CITRA MAMMOGRAM BERDASARKAN FITUR BENTUK

• Main Authors: , Miftahus Sholihin, , Drs. Agus Harjoko, M.Sc., Ph.D.
• Format: Theses and Dissertations NonPeerReviewed
• Published: [Yogyakarta] : Universitas Gadjah Mada 2013
• Subjects: ETD
• Online Access: https://repository.ugm.ac.id/126177/; http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=66375
• Institution: Universitas Gadjah Mada

Breast cancer is one of the most dangerous types of cancer that affects women all over the world. Based on characteristic mammography image a physician or radiologist can find change in the breast. Analysis of mammography image performed by radiologist still done manually. So the result of the analysis of radiological or doctor are subjective, in addition to the results of the observation of doctor sometimes take a long time and also the result obtained sometimes not appropriate. Therefore we need a tool that can accelerate radiologist to performance using digital image processing techniques. In this study desined a system that can help the radiologist in the classification of cancer on the mammogram image. This process begins by elimination of the background and label image, then segmentation of pectoral muscle, the next process is to improving the quality of the image, the next process is segmentation of suspected area cancer, the next process is feature extraction using invariant moment and the final process is to classification using KNN method. This process will determine the classification of mammogram image into normal classes, benign, and malignant. Data used in this study take from MIAS database totaling 126 image consisting of 40 normal, 48 benign, and 38 malignant. Accuracy obtained 76,9% for class normal and abnormal when the value of k = 9. While the average accuracy obtained amounted 50,39% to classification into class benign and class malignant.