Paper Title
Quantification of HER2 Immunohistochemistry Breast Cancer Patients using K-Mean Clustering

Abstract
Patient was diagnosed with cancer, it is necessary to do a series of tests such as hormone receptor status of the affected tissue. The type of test that tests the gene HER2 (human epidermal growth factor receptor-2). This gene associated with aggressive cancer cell growth. The patient is said to be HER2-positive if the tumor is HER2 is found in large quantities. To be able to display the immunohistochemistry test in numerical magnitude required software in the form of medical image analysis. With this application checks can describe the intensity of staining in immunohistochemistry examination as numeric variables are more appropriate than qualitative visual observation and can display the percentage of cells. The aim of this research is to develop a medical image analysis application that is able to quantify the results of immunohistochemistry examinations. By using k Mean Clustering so that the segmentation process becomes easier. By starting the process morphology, segmentation is done by using K-Mean clustering techniques to obtain the extent of the cell nucleus and cytoplasm area. To test the grade totals obtained the best accuracy of 94.44 % in the number of grade 10. The test using immunohistochemistry images with HER2 staining obtained recognition percentage rate of 91.72. Keywords - Her2, Immunohistochemistry, Breast Cancer, K-Mean Clustering