Home // BIOTECHNO 2011, The Third International Conference on Bioinformatics, Biocomputational Systems and Biotechnologies // View article
Authors:
Lee Sing Cheong
Angela Jean
Tsu Soo Tan
Waiming Kong
Soo Yong Tan
Keywords: HER2; neu; breast carcinoma; automated segmentation and measurement; cancer classification.
Abstract:
The HER2/neu protein is over-expressed in 20%-30% of breast cancer cases, and is significantly associated with increased breast cancer recurrence and worse prognosis. The assessment of HER2 protein level is visualized using immunohistochemistry (IHC) assays, which is subjected to inter-observer and intra-observer variability. In this paper, we reduce variability by an automated segmentation and measurement system for IHC-stained breast tissue images. From the dataset of breast tissue images, the system is able to obtain the nuclei and the orange stained cell membranes of the cells, quantify the continuous orange hue of the cell membranes, and identify nuclei that are bounded by orange stained cell membrane. This system also suggests a putative assessment classification score for each image based on the same assessment protocol specified for histopathologist. Using the dataset of 42 clinically IHC-scored images, the system correctly suggested the corresponding putative assessment classification score for 39 of the images, achieving an accuracy of 92%.
Pages: 43 to 48
Copyright: Copyright (c) IARIA, 2011
Publication date: May 22, 2011
Published in: conference
ISSN: 2308-4383
ISBN: 978-1-61208-137-3
Location: Venice/Mestre, Italy
Dates: from May 22, 2011 to May 27, 2011