Decision based algorithm for gene markers detection in the ISH images

Abstract

The article presents a decision based algorithm of an automatic identification of markers in microscope images—ISH (In Situ Hybridization). The ISH test allows a quick and inexpensive initial diagnosis of the breast cancer. The evaluation of a degree of the HER2 gene’s amplification and the selection of the appropriate treatment require locating and counting markers in cell nuclei. This article presents a new heterogeneous algorithm based on decision making. The main idea is to analyze a portion of an image and decide which algorithm should be used for processing the given fragment. The different parts of the image can be analyzed by different types of algorithms. Tests and results of the experiment are presented and discussed in this article. They confirm higher efficiency of markers recognition by the homogeneous system, using many different algorithms, rather than in the case of using the method based on a single algorithm.