Klasifikasi Jenis Kanker Payudara Menggunakan K-Neighbor Dengan Fitur Gray Level Co-Occurrence Matrix Dan Fitur Dispersi

Abstract

Breast most cancers is a ailment that takes place because of the boom or improvement of breast cells (tissue), this may arise in female and male. This sickness has a reasonably excessive prevalence fee in evolved countries. Breast cancer results from the abnormal and uncontrolled growth of cells in the breast, typically forming a lump-like tumor. There are several ways that can be done to check for breast cancer including self-examination or better known as BSE (Self-Breast Examination) and examinations with medical assistance which are often done with Magnetic Resonance Imaging (MRI), X-ray mammograms and ultrasounds (USG). However, the resulting image or image still has noise. Image processing can be used to identify images into objects based on certain characteristics. To be able to identify the image, you can use the GLCM method and use the Dispersion method and to do the classification using the K-Nearest Neighboor method. Testing using 30 ultrasound images of breast cancer which are divided to 20 training data and 10 test data. And each in the data is divided into types of benign breast cancer and types of malignant breast cancer. The results of this study achieved an accuracy of 80%. This study uses the MATLAB R2017b application.