Segmentation of breast cancer cells positive 1+ and 3+ immunohistochemistry

Breast cancer is a disease occurs as a result of uncontrolled cells growth. One examination method of breast cancer cells is using Immunohistochemistry (IHC) to determine status of Human Epidermal Growth Factor Receptor2 (HER2) protein. This study helps anatomic pathologist to determine HER2 scores using image processing techniques to obtain HER2 overexpression positive area percentages of 1+ and 3+ scores. This is done because the score of 0 is HER2 negative cells and 2+ scores have equivocal results, which means it could not be determined whether it is necessary to give targeted therapy or not. HER2 overexpression positive area percentage is done by dividing the area with a HER2 positive tumor area. To obtain better tumor area, repair is done by eliminating lymphocytes area which is not tumor area using morphological opening. Results of 10 images IHC scores of 1+ and 3+ and 10 IHC images testing without losing lymphocytes area in tumor area, has proven that the system has been able to provide an overall correct classification in accordance with the experts analysis. However by doing operation to remove non-tumor areas, classification can be done correctly 100% for scores of 3+ and 65% for scores of 1+.