HEp-2 cell pattern segmentation for the support of autoimmune disease diagnosis

Abstract

The Indirect Immune Fluorescence Test (iIFT) is the most commonly used screening method for the diagnosis of autoimmune diseases. The presence of certain autoimmune diseases is proven by immunologically detecting their corresponding auto-antibodies using the HEp-2 cancer cell line. For this purpose HEp-2 cells are added to the patients' blood serum containing certain auto-antibodies which will bond with the HEp-2 cells leading to a wide variety of patterns that can be observed under a fluorescence microscope. Due to the disadvantages of manual testing, automation and standardization are necessary. This paper proposes an unsupervised segmentation algorithm as part of an ongoing research to develop a CAD system to digitally support iIFT testing.