A user-friendly tool for cloud-based whole slide image segmentation with examples from renal histopathology
Brendon Lutnick, David Manthey, Jan U. Becker, Brandon Ginley, Katharina Moos, Jonathan E. Zuckerman, Luís Rodrigues, Alexander J. Gallan, Laura Barisoni, Charles E. Alpers, Xiaoxin X. Wang, Komuraiah Myakala, Bryce A. Jones, Moshe Levi, Jeffrey B. Kopp, Teruhiko Yoshida, Jarcy Zee, Seung Seok Han, Sanjay Jain, Avi Z. Rosenberg, Kuang‐Yu Jen, Pinaki Sarder, the Kidney Precision Medicine Project, Brendon Lutnick, Brandon Ginley, Richard Knight, Stewart H. Lecker, Isaac E. Stillman, Steve Bogen, Afolarin Amodu, Titlayo Ilori, Insa M. Schmidt, Shana Maikhor, Laurence H. Beck, Ashish Verma, Joel Henderson, Ingrid Onul, Sushrut S. Waikar, Gearoid M. McMahon, Astrid Weins, Mia R. Colona, M. Todd Valerius, Nir Hacohen, Paul Hoover, Anna Greka, Jamie L. Marshall, Mark P. Aulisio, Yijiang M. Chen, Andrew Janowczyk, Catherine Jayapandian, Vidya Sankar Viswanathan, William S. Bush, Dana C. Crawford, Anant Madabhushi, John O’Toole, Emilio D. Poggio, John R. Sedor, Leslie Cooperman, Stacey E. Jolly, Leal Herlitz, Jane Nguyen, Agustin Gonzalez‐Vicente, Ellen L. Palmer, Dianna Sendrey, Jonathan J. Taliercio, Lakeshia Bush, Kassandra Spates-Harden, Carissa Vinovskis, P. M. Bjørnstad, Laura Pyle, Paul S. Appelbaum, Jonathan Barasch, Andrew S. Bomback, Vivette D. D’Agati, Krzysztof Kiryluk, Karla Mehl, Pietro A. Canetta, Ning Shang, Olivia Balderes, Satoru Kudose, Theodore Alexandrov, Helmut G. Rennke, Tarek M. El‐Achkar, Ying‐Hua Cheng, Pierre C. Dagher, Michael T. Eadon, Kenneth W. Dunn, Katherine J. Kelly, Timothy A. Sutton, Daria Barwinska, Michael J. Ferkowicz, Seth Winfree, Sharon B. Bledsoe, Marcelino Rivera, James C. Williams, Ricardo Melo Ferreira, Katy Börner, Andreas Bueckle, Bruce Herr, Ellen M. Quardokus
Abstract
Background: Image-based machine learning tools hold great promise for clinical applications in pathology research. However, the ideal end-users of these computational tools (e.g., pathologists and biological scientists) often lack the programming experience required for the setup and use of these tools which often rely on the use of command line interfaces. Methods: , a tool for segmentation of whole slide images (WSIs) that has an easy-to-use graphical user interface. This tool runs a state-of-the-art convolutional neural network (CNN) for segmentation of WSIs in the cloud and allows the extraction of features from segmented regions for further analysis. Results: By segmenting glomeruli, interstitial fibrosis and tubular atrophy, and vascular structures from renal and non-renal WSIs, we demonstrate the scalability, best practices for transfer learning, and effects of dataset variability. Finally, we demonstrate an application for animal model research, analyzing glomerular features in three murine models. Conclusions: is open source, accessible over the internet, and adaptable for segmentation of any histological structure regardless of stain.