Efficacy of deep convolutional neural network features on histological manifold for robust breast carcinoma detection

Subhankar Chattoraj; Souvik Pratiher; Rajdeep Mukherjee; Saikat Ghosh; Arnab Chakraborty; Diptiman Hazra; Sawon Pratiher

doi:10.1117/12.2505522

1 November 2018 Efficacy of deep convolutional neural network features on histological manifold for robust breast carcinoma detection

Subhankar Chattoraj, Souvik Pratiher, Rajdeep Mukherjee, Saikat Ghosh, Arnab Chakraborty, Diptiman Hazra, Sawon Pratiher

Author Affiliations +

Proceedings Volume 10820, Optics in Health Care and Biomedical Optics VIII; 108203Q (2018) https://doi.org/10.1117/12.2505522
Event: SPIE/COS Photonics Asia, 2018, Beijing, China

Abstract

In this study, exploratory deep feature engineering using convolutional neural network (CNN) on histological manifold has been proposed for robust breast carcinoma detection. A comparative evaluation emphasizing the adequacy of manifold learning and CNN aided deep features over state-of-the-art biomarkers and other deep learning models is done for histopathological image (HI) classification. The proposed framework efficiently differentiate the spatial textural non-stationarity in HI and apprehend the topographic aberrations of cancerous tissues and exemplifies its competency for clinical settings deployment in developing countries. Experimental results are discussed in detail.

Citation Download Citation

Subhankar Chattoraj, Souvik Pratiher, Rajdeep Mukherjee, Saikat Ghosh, Arnab Chakraborty, Diptiman Hazra, and Sawon Pratiher "Efficacy of deep convolutional neural network features on histological manifold for robust breast carcinoma detection", Proc. SPIE 10820, Optics in Health Care and Biomedical Optics VIII, 108203Q (1 November 2018); https://doi.org/10.1117/12.2505522

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available