Skip to main content

Research Repository

Advanced Search

Dual Adaptive Pyramid Network for Cross-Stain Histopathology Image Segmentation

Hou, Xianxu; Liu, Jingxin; Xu, Bolei; Liu, Bozhi; Chen, Xin; Garibaldi, Jon; Ilyas, Mohammad; Ellis, Ian; Qiu, Guoping

Dual Adaptive Pyramid Network for Cross-Stain Histopathology Image Segmentation Thumbnail


Authors

Xianxu Hou

Jingxin Liu

Bolei Xu

Bozhi Liu

XIN CHEN XIN.CHEN@NOTTINGHAM.AC.UK
Associate Professor

Jon Garibaldi

GUOPING QIU GUOPING.QIU@NOTTINGHAM.AC.UK
Professor of Visual Information Processing



Abstract

Supervised semantic segmentation normally assumes the test data being in a similar data domain as the training data. However, in practice, the domain mismatch between the training and unseen data could lead to a significant performance drop. Obtaining accurate pixel-wise label for images in different domains is tedious and labor intensive, especially for histopathology images. In this paper, we propose a dual adaptive pyramid network (DAPNet) for histopathological gland segmentation adapting from one stain domain to another. We tackle the domain adaptation problem on two levels: (1) the image-level considers the differences of image color and style; (2) the feature-level addresses the spatial inconsistency between two domains. The two components are implemented as domain classifiers with adversarial training. We evaluate our new approach using two gland segmentation datasets with H&E and DAB-H stains respectively. The extensive experiments and ablation study demonstrate the effectiveness of our approach on the domain adaptive segmentation task. We show that the proposed approach performs favorably against other state-of-the-art methods.

Online Publication Date Oct 10, 2019
Publication Date 2019
Deposit Date Jan 10, 2020
Publicly Available Date Oct 11, 2020
Publisher Springer Verlag
Pages 101-109
Series Title Lecture Notes in Computer Science
Series Number 11765
Book Title Medical Image Computing and Computer Assisted Intervention – MICCAI 2019: 22nd International Conference, Shenzhen, China, October 13–17, 2019, Proceedings, Part II
ISBN 9783030322441
DOI https://doi.org/10.1007/978-3-030-32245-8_12
Public URL https://nottingham-repository.worktribe.com/output/3518035
Publisher URL https://link.springer.com/chapter/10.1007%2F978-3-030-32245-8_12
Additional Information First Online: 10 October 2019; Conference Acronym: MICCAI; Conference Name: International Conference on Medical Image Computing and Computer-Assisted Intervention; Conference City: Shenzhen; Conference Country: China; Conference Year: 2019; Conference Start Date: 13 October 2019; Conference End Date: 17 October 2019; Conference Number: 22; Conference ID: miccai2019; Conference URL: https://www.miccai2019.org/; Type: Double-blind; Conference Management System: CMT; Number of Submissions Sent for Review: 1730; Number of Full Papers Accepted: 539; Number of Short Papers Accepted: 0; Acceptance Rate of Full Papers: 31% - The value is computed by the equation "Number of Full Papers Accepted / Number of Submissions Sent for Review * 100" and then rounded to a whole number.; Average Number of Reviews per Paper: 3.07; Average Number of Papers per Reviewer: 6.31; External Reviewers Involved: Yes

Files





You might also like



Downloadable Citations