Automatic evaluation system of FISH images in breast cancer Tomasz Les 1 , Tomasz Markiewicz 1,2 , Stanisław Osowski 1,3 , Marzena Cichowicz 2 , Wojciech Kozlowski 2 1 Warsaw University of Technology, 2 Military Institute of Medicine, 3 Military University of Technology, Warsaw, Poland lestomasz@gmail.com, {markiewt,sto}@iem.pw.edu.pl, {mcicho- wicz, wkozlowski}@wim.mil.pl Abstract. The paper presents the algorithm of an automatic evaluation of the fluorescent in situ hybridization (FISH) images in order to determine HER2 status of the breast cancer samples. The algorithm is based on the accurate measurement of the red/green spot ratio (the ratio of HER2/CEN17) per cell nucleus. The main points of this algorithm is an accurate detection of the nuclei of cells and applica- tion of the color map to detect the red and green spots localized in these cells. The results of the numerical experiments concerning these two problems, as well as the assessment of the positivity or negativity of the considered cases are presented and discussed in the paper. They confirm the high efficiency of the proposed solu- tion. Key words: image segmentation, object recognition, FISH images 1 Introduction The fluorescence in situ hybridization (FISH) test performed on the breast cancer tis- sue removed during the biopsy allows to visualize the specific genes to discover if extra copies of the HER2 gene have been created. More copies of the HER2 gene mean that more HER2 receptors the cells have, which stimulate the growth of the breast cancer cells. The probes carry the fluorescent markers emitting the light after binding to the HER2 genes. These HER2 probes are visible as the red/orange stained spots under a fluorescent microscope. On the other hand the probes for the chromosome centromere (CEN17) on which HER2 is located are visible as the green spots. The ratio between the number of the red spots over the number of green spots estimated for the investi- gated case, decides whether the cancer is positive or negative for HER2 and decides of the way of the medical treatment. Inaccurate results of HER2 test may cause appli- cation of the not accurate diagnosis and as a result the not optimal treatment. The manual analysis of the FISH images is a time consuming and prone to er- rors, since it involves the manual counting of hundreds of the FISH spots over the