| Literature DB >> 30364524 |
Elliot Gray1, Elizabeth Mitchell2, Sonali Jindal2, Pepper Schedin2, Young Hwan Chang1.
Abstract
Ductal carcinoma in situ (DCIS) is breast cancer confined within mammary ducts, surrounded by an intact myoepithelial cell layer that prevents local invasion. A DCIS diagnosis confers increased lifetime risk of developing invasive breast cancer (IBC) and results in surgical excision with radiation, and possibly endocrine- or chemo-therapy. DCIS is known to be over treated, with associated co-morbidities. Biomarkers are needed that delineate patients at low risk of DCIS progression from patients requiring more aggressive treatment. Investigating the role of myoepithelial cell differentiation in barrier function is anticipated to provide insight into DCIS progression and delineate between low and high risk lesions. Here, we develop a high throughput technique to assess loss of myoepithelial differentiation markers. This method facilitates automated analysis of a clinically relevant histopathologic feature, as demonstrated by a high correlation with pathologist annotation (r = 0.959), and further, contributes analytical foundations to a multiplexed immunohistochemistry (IHC) approach.Entities:
Keywords: DCIS; calponin; feature engineering; invasive breast cancer; myoepithelial cell
Year: 2018 PMID: 30364524 PMCID: PMC6196724 DOI: 10.1109/ISBI.2018.8363692
Source DB: PubMed Journal: Proc IEEE Int Symp Biomed Imaging ISSN: 1945-7928