Title :
Automated Her2 Assessment of FISH Images with Color Compensation
Author :
Kuo, Yung-Lung ; Ko, Chien-Chuan ; Lai, Jen-Yuan
Author_Institution :
Dept. of Comput. Sci. & Inf. Eng., Far East Univ., Tainan, Taiwan
Abstract :
HER2 stands for human epidermal growth receptor 2 is a receptor tyrosine kinase. It hands in the signal transduction pathway about cell growing and differentiation. HER2 gen amplification status can be assessed with fluorescence in situ hybridization by computing the ratio of HER2 over Centro mere 17 markers detected within a number of representative nuclei. However, owing to florescent spreading, color compensation is necessary. In this article, we present an automated model for the quantitative assessment of HER2 after compensating color of microscopy image. The methods processed the tissue with a multistage algorithm, including signals segmentation, nuclei segmentation and reliable scoring. And experimental results show the agreement of HER-2 amplification between the proposed scheme and expert pathologist.
Keywords :
image colour analysis; image segmentation; medical image processing; FISH images; automated Her2 assessment; color compensation; human epidermal growth receptor 2; microscopy image; multistage algorithm; nuclei segmentation; receptor tyrosine kinase; reliable scoring; representative nuclei; signal transduction pathway; signals segmentation; Accuracy; Biological cells; Breast cancer; Fluorescence; Image color analysis; Image segmentation; Marine animals; color compensation; fluorescence in-situ hybridization; nuclei segmentation; watershed segmentation;
Conference_Titel :
Computer, Consumer and Control (IS3C), 2012 International Symposium on
Conference_Location :
Taichung
Print_ISBN :
978-1-4673-0767-3
DOI :
10.1109/IS3C.2012.210