Whole-Slide Mitosis Detection in H&E Breast Histology Using PHH3 as a Reference to Train Distilled Stain-Invariant Convolutional Networks
David Tellez,Maschenka Balkenhol,Irene Otte-Höller,Rob van de Loo,Rob Vogels,Peter Bult,Carla Wauters,Willem Vreuls,Suzanne Mol,Nico Karssemeijer,Geert Litjens,Jeroen van der Laak,Francesco Ciompi +12 more
Reads0
Chats0
TLDR
In this paper, a method to automatically detect mitotic tumor cells in breast cancer tissue sections based on convolutional neural networks (CNNs) was developed, which was trained in a single-center cohort and evaluated in an independent multicenter cohort from the cancer genome atlas.Abstract:
Manual counting of mitotic tumor cells in tissue sections constitutes one of the strongest prognostic markers for breast cancer. This procedure, however, is time-consuming and error-prone. We developed a method to automatically detect mitotic figures in breast cancer tissue sections based on convolutional neural networks (CNNs). Application of CNNs to hematoxylin and eosin (H&E) stained histological tissue sections is hampered by noisy and expensive reference standards established by pathologists, lack of generalization due to staining variation across laboratories, and high computational requirements needed to process gigapixel whole-slide images (WSIs). In this paper, we present a method to train and evaluate CNNs to specifically solve these issues in the context of mitosis detection in breast cancer WSIs. First, by combining image analysis of mitotic activity in phosphohistone-H3 restained slides and registration, we built a reference standard for mitosis detection in entire H&E WSIs requiring minimal manual annotation effort. Second, we designed a data augmentation strategy that creates diverse and realistic H&E stain variations by modifying H&E color channels directly. Using it during training combined with network ensembling resulted in a stain invariant mitosis detector. Third, we applied knowledge distillation to reduce the computational requirements of the mitosis detection ensemble with a negligible loss of performance. The system was trained in a single-center cohort and evaluated in an independent multicenter cohort from the cancer genome atlas on the three tasks of the tumor proliferation assessment challenge. We obtained a performance within the top three best methods for most of the tasks of the challenge.read more
Citations
More filters
Posted Content
WILDS: A Benchmark of in-the-Wild Distribution Shifts
Pang Wei Koh,Shiori Sagawa,Henrik Marklund,Sang Michael Xie,Marvin Zhang,Akshay Balsubramani,Weihua Hu,Michihiro Yasunaga,Richard Lanas Phillips,Irena Gao,Tony Lee,Etienne David,Ian Stavness,Wei Guo,Berton A. Earnshaw,Imran S. Haque,Sara Beery,Jure Leskovec,Anshul Kundaje,Emma Pierson,Sergey Levine,Chelsea Finn,Percy Liang +22 more
TL;DR: WILDS is presented, a benchmark of in-the-wild distribution shifts spanning diverse data modalities and applications, and is hoped to encourage the development of general-purpose methods that are anchored to real-world distribution shifts and that work well across different applications and problem settings.
Journal ArticleDOI
Quantifying the effects of data augmentation and stain color normalization in convolutional neural networks for computational pathology.
David Tellez,Geert Litjens,Péter Bándi,Wouter Bulten,John-Melle Bokhorst,Francesco Ciompi,Jeroen van der Laak +6 more
TL;DR: In this article, the authors compared stain color augmentation and normalization techniques and quantified their effect on CNN classification performance using a heterogeneous dataset of hematoxylin and eosin histopathology images from 4 organs and 9 pathology laboratories.
Journal ArticleDOI
Deep neural network models for computational histopathology: A survey
TL;DR: A comprehensive review of state-of-the-art deep learning approaches that have been used in the context of histopathological image analysis can be found in this paper, where a survey of over 130 papers is presented.
Journal ArticleDOI
Computational pathology definitions, best practices, and recommendations for regulatory guidance: a white paper from the Digital Pathology Association.
Esther Abels,Liron Pantanowitz,Famke Aeffner,Mark D. Zarella,Jeroen van der Laak,Jeroen van der Laak,Marilyn M. Bui,Venkata N. P. Vemuri,Anil V. Parwani,Jeff Gibbs,Emmanuel Agosto-Arroyo,Andrew H. Beck,Cleopatra Kozlowski +12 more
TL;DR: In this paper, the authors define terminology and concepts in the emerging field of computational pathology, with a focus on its application to histology images analyzed together with their associated patient data to extract information.
Journal ArticleDOI
Deep learning in histopathology: the path to the clinic
TL;DR: In the field of histopathology, deep learning algorithms have been developed that perform similarly to trained pathologists for tasks such as tumor detection and grading, but despite these promising results, very few algorithms have reached clinical implementation, challenging the balance between hope and hype for these new techniques as discussed by the authors.
References
More filters
Proceedings ArticleDOI
ImageNet: A large-scale hierarchical image database
TL;DR: A new database called “ImageNet” is introduced, a large-scale ontology of images built upon the backbone of the WordNet structure, much larger in scale and diversity and much more accurate than the current image datasets.
Dissertation
Learning Multiple Layers of Features from Tiny Images
TL;DR: In this paper, the authors describe how to train a multi-layer generative model of natural images, using a dataset of millions of tiny colour images, described in the next section.
Posted Content
Distilling the Knowledge in a Neural Network
TL;DR: This work shows that it can significantly improve the acoustic model of a heavily used commercial system by distilling the knowledge in an ensemble of models into a single model and introduces a new type of ensemble composed of one or more full models and many specialist models which learn to distinguish fine-grained classes that the full models confuse.
Journal ArticleDOI
A survey on deep learning in medical image analysis
Geert Litjens,Thijs Kooi,Babak Ehteshami Bejnordi,Arnaud Arindra Adiyoso Setio,Francesco Ciompi,Mohsen Ghafoorian,Jeroen van der Laak,Bram van Ginneken,Clara I. Sánchez +8 more
TL;DR: This paper reviews the major deep learning concepts pertinent to medical image analysis and summarizes over 300 contributions to the field, most of which appeared in the last year, to survey the use of deep learning for image classification, object detection, segmentation, registration, and other tasks.
Journal ArticleDOI
The cancer genome atlas pan-cancer analysis project
John N. Weinstein,John N. Weinstein,Eric A. Collisson,Gordon B. Mills,Kenna R. Mills Shaw,Kenna R. Mills Shaw,Brad Ozenberger,Kyle Ellrott,Kyle Ellrott,Chris Sander,Joshua M. Stuart,Joshua M. Stuart,Kyle Chang,Chad J. Creighton,Caleb F. Davis,Lawrence A. Donehower,Jennifer Drummond,David A. Wheeler,Adrian Ally,Miruna Balasundaram,Inanc Birol,Inanc Birol,Inanc Birol,Yaron S.N. Butterfield,Andy Chu,Eric Chuah,Hye Jung E. Chun,Noreen Dhalla,Ranabir Guin,Martin Hirst,Carrie Hirst,Robert A. Holt,Steven J.M. Jones,Darlene Lee,Haiyan I. Li,Marco A. Marra,Michael Mayo,Richard A. Moore,Andrew J. Mungall,A. Gordon Robertson,Jacqueline E. Schein,Payal Sipahimalani,Angela Tam,Nina Thiessen,Richard Varhol,Rameen Beroukhim,Ami S. Bhatt,Angela N. Brooks,Andrew D. Cherniack,Samuel S. Freeman,Stacey Gabriel,Elena Helman,Joonil Jung,Matthew Meyerson,Akinyemi I. Ojesina,Chandra Sekhar Pedamallu,Gordon Saksena,Steven E. Schumacher,Barbara Tabak,Travis I. Zack,Travis I. Zack,Eric S. Lander,Christopher A. Bristow,Angela Hadjipanayis,Psalm Haseley,Raju Kucherlapati,Semin Lee,Eunjung Lee,Lovelace J. Luquette,Harshad S. Mahadeshwar,Angeliki Pantazi,Michael Parfenov,Michael Parfenov,Peter J. Park,Alexei Protopopov,Xiaojia Ren,Netty Santoso,Jonathan G. Seidman,Sahil Seth,Xingzhi Song,Jiabin Tang,Ruibin Xi,Ruibin Xi,Ruibin Xi,Andrew Wei Xu,Lixing Yang,Dong Zeng,J. Todd Auman,Saianand Balu,Elizabeth Buda,Cheng Fan,Katherine A. Hoadley,Corbin D. Jones,Shaowu Meng,Piotr A. Mieczkowski,Joel S. Parker,Charles M. Perou,Jeffrey Roach,Yan Shi,Grace O. Silva,Donghui Tan,Umadevi Veluvolu,Scot Waring,Matthew D. Wilkerson,Junyuan Wu,Wei Zhao,Tom Bodenheimer,D. Neil Hayes,D. Neil Hayes,Alan P. Hoyle,Stuart R. Jeffreys,Lisle E. Mose,Janae V. Simons,Mathew G. Soloway,Stephen B. Baylin,Benjamin P. Berman,Moiz S. Bootwalla,Ludmila Danilova,James G. Herman,Toshinori Hinoue,Peter W. Laird,Suhn K. Rhie,Hui Shen,Timothy J. Triche,Daniel J. Weisenberger,Scott L. Carter,Kristian Cibulskis,Lynda Chin,Jianhua Zhang,Carrie Sougnez,Min Wang,Gad Getz,Gad Getz,Huyen Dinh,Harshavardhan Doddapaneni,Richard A. Gibbs,Preethi Gunaratne,Preethi Gunaratne,Yi Han,Divya Kalra,Christie Kovar,Lora Lewis,Margaret B. Morgan,Donna Morton,Donna Muzny,Jeffrey G. Reid,Liu Xi,Juok Cho,Daniel DiCara,Scott Frazer,Nils Gehlenborg,David I. Heiman,Jaegil Kim,Michael S. Lawrence,Pei Lin,Yingchun Liu,Michael S. Noble,Petar Stojanov,Doug Voet,Hailei Zhang,Lihua Zou,Chip Stewart,Brady Bernard,Ryan Bressler,Andrea Eakin,Lisa Iype,Theo A. Knijnenburg,Roger Kramer,Richard Kreisberg,Kalle Leinonen,Jake Lin,Yuexin Liu,Michael Miller,Sheila M. Reynolds,Hector Rovira,Ilya Shmulevich,Vesteinn Thorsson,Da Yang,Wei Zhang,Samirkumar B. Amin,Chang-Jiun Wu,Chia Chin Wu,Rehan Akbani,Kenneth Aldape,Keith A. Baggerly,Bradley McIntosh Broom,Tod D. Casasent,James Cleland,Deepti Dodda,Mary Elizabeth Edgerton,Leng Han,Shelley M. Herbrich,Zhenlin Ju,Hoon Kim,Hoon Kim,Seth Lerner,Jun Li,Han Liang,Wenbin Liu,Philip L. Lorenzi,Yiling Lu,James M. Melott,Lam Nguyen,Lam Nguyen,Xiaoping Su,Roeland Verhaak,Wenyi Wang,Andrew J. Wong,Andrew J. Wong,Yang Yang,Jun Yao,Rong Yao,Kosuke Yoshihara,Yuan Yuan,Yuan Yuan,W. K. Alfred Yung,Nianxiang Zhang,Siyuan Zheng,Michael B. Ryan,Michael B. Ryan,David W. Kane,David W. Kane,B. Arman Aksoy,Giovanni Ciriello,Gideon Dresdner,Jianjiong Gao,Benjamin Gross,Anders Jacobsen,André Kahles,Marc Ladanyi,William Lee,Kjong-Van Lehmann,Martin L. Miller,Ricardo Ramirez,Gunnar Rätsch,Boris Reva,Nikolaus Schultz,Yasin Senbabaoglu,Ronglai Shen,Rileen Sinha,S. Onur Sumer,Yichao Sun,Barry S. Taylor,Barry S. Taylor,Barry S. Taylor,Nils Weinhold,Suzanne S. Fei,Paul T. Spellman,Christopher C. Benz,Christopher C. Benz,Daniel E. Carlin,Daniel E. Carlin,Melisssa Cline,Melisssa Cline,Brian Craft,Brian Craft,Mary Goldman,David Haussler,David Haussler,David Haussler,Singer Ma,Singer Ma,Sam Ng,Sam Ng,Evan O. Paull,Evan O. Paull,Amie Radenbaugh,Amie Radenbaugh,Sofie R. Salama,Sofie R. Salama,Sofie R. Salama,Artem Sokolov,Artem Sokolov,Teresa Swatloski,Teresa Swatloski,Vladislav Uzunangelov,Vladislav Uzunangelov,Peter Waltman,Peter Waltman,Christina Yau,Jing Zhu,Jing Zhu,Stanley R. Hamilton,Scott Abbott,Rachel Abbott,Nathan D. Dees,Kim D. Delehaunty,Li Ding,David J. Dooling,James M. Eldred,Catrina Fronick,Robert S. Fulton,Lucinda Fulton,Joelle Kalicki-Veizer,Krishna L. Kanchi,Cyriac Kandoth,Daniel C. Koboldt,David E. Larson,Timothy J. Ley,Ling Lin,Charles Lu,Vincent Magrini,Elaine R. Mardis,Michael D. McLellan,Joshua F. McMichael,Christopher A. Miller,Michelle O'Laughlin,Craig Pohl,Heather Schmidt,Scott M. Smith,Jason Walker,John W. Wallis,Michael C. Wendl,Michael C. Wendl,Richard K. Wilson,Todd Wylie,Qunyuan Zhang,Robert A. Burton,Mark A. Jensen,Ari B. Kahn,Todd Pihl,David A. Pot,Yunhu Wan,Douglas A. Levine,Aaron D. Black,Jay Bowen,Jessica Frick,Julie M. Gastier-Foster,Julie M. Gastier-Foster,Hollie A. Harper,Carmen Helsel,Kristen M. Leraas,Tara M. Lichtenberg,Cynthia McAllister,Nilsa C. Ramirez,Nilsa C. Ramirez,Samantha Sharpe,Lisa Wise,Erik Zmuda,Stephen J. Chanock,Tanja Davidsen,John A. Demchok,Greg Eley,Ina Felau,Margi Sheth,Heidi J. Sofia,Louis M. Staudt,Roy Tarnuzzer,Zhining Wang,Liming Yang,Jiashan Zhang,Larsson Omberg,Adam Margolin,Benjamin J. Raphael,Fabio Vandin,Hsin-Ta Wu,Mark D.M. Leiserson,Stephen C. Benz,Charles J. Vaske,Houtan Noushmehr,Houtan Noushmehr,Denise M. Wolf,Laura van 't Veer,Dimitris Anastassiou,Tai Hsien Ou Yang,Nuria Lopez-Bigas,Abel Gonzalez-Perez,David Tamborero,Zheng Xia,Wei Li,Dong Yeon Cho,Teresa M. Przytycka,Mark P. Hamilton,Sean E. McGuire,Sven Nelander,Sven Nelander,Patrik Johansson,Rebecka Jörnsten,Rebecka Jörnsten,Teresia Kling +379 more
TL;DR: The Pan-Cancer initiative compares the first 12 tumor types profiled by TCGA with a major opportunity to develop an integrated picture of commonalities, differences and emergent themes across tumor lineages.
Related Papers (5)
Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node Metastases in Women With Breast Cancer.
Babak Ehteshami Bejnordi,Mitko Veta,Paul J. van Diest,Bram van Ginneken,Nico Karssemeijer,Geert Litjens,Jeroen van der Laak,Meyke Hermsen,Quirine F. Manson,Maschenka Balkenhol,Oscar Geessink,N. Stathonikos,Marcory C. R. F. van Dijk,Peter Bult,Francisco Beca,Andrew H. Beck,Dayong Wang,Aditya Khosla,Rishab Gargeya,Humayun Irshad,Aoxiao Zhong,Qi Dou,Qi Dou,Quanzheng Li,Hao Chen,Huangjing Lin,Pheng-Ann Heng,Christian Haß,Elia Bruni,Quincy Wong,Ugur Halici,Mustafa Umit Oner,Rengul Cetin-Atalay,Matt Berseth,Vitali Khvatkov,Alexei Vylegzhanin,Oren Kraus,Muhammad Shaban,Nasir M. Rajpoot,Nasir M. Rajpoot,Ruqayya Awan,Korsuk Sirinukunwattana,Talha Qaiser,Yee-Wah Tsang,David Tellez,Jonas Annuscheit,Peter Hufnagl,Mira Valkonen,Kimmo Kartasalo,Kimmo Kartasalo,Leena Latonen,Pekka Ruusuvuori,Pekka Ruusuvuori,Kaisa Liimatainen,Shadi Albarqouni,Bharti Mungal,Ami George,Stefanie Demirci,Nassir Navab,Seiryo Watanabe,Shigeto Seno,Yoichi Takenaka,Hideo Matsuda,Hady Ahmady Phoulady,Vassili Kovalev,A. Kalinovsky,Vitali Liauchuk,Gloria Bueno,M. Milagro Fernández-Carrobles,Ismael Serrano,Oscar Deniz,Daniel Racoceanu,Daniel Racoceanu,Rui Venâncio +73 more