THE SYSTEM FOR PATHOMORPHOLOGICAL MICROSCOPIC IMAGE ANALYSIS

We present a computer system supporting the analysis of microscopic slides. Our goal was to automate the laborious process of counting the cell nuclei in a viewing area and to indicate their type basing on immunohistochemical reaction. To prove the reliability of the system we do a statistical comparison of our results to the results obtained manually by a qualified lab personnel. 1. PATHOMORPHOLOGICAL EXAMINATION The aim of the pathomorphological examination that we intend to automate [2] is detection of estrogen and progesterone receptors in cancer tissue nuclei, which are responsible for tissue growth. One of the detection techniques is immunohistochemical coloring of the tissue. Nuclei with receptors react with a dye and change colors into red-brown, while other remain pale blue. The fraction of colored nuclei in the specific area gives an important hint on choosing proper pharmacological therapy. In order to perform a routine examination, we need cancer tissue material of four or five microscopic slides: two from biopsy, two from histological extracts, and one from a frozen plate imprint