Domain adaptation model for retinopathy detection from cross-domain OCT images

Jing Wang, Yiwei Chen, Wanyue Li, Wen Kong, Yi He, Chuihui Jiang, Guohua Shi
; Proceedings of the Third Conference on Medical Imaging with Deep Learning, PMLR 121:795-810, 2020.

Abstract

A deep neural network (DNN) can assist in retinopathy screening by automatically classifying patients into normal and abnormal categories according to optical coherence tomography (OCT) images. Typically, OCT images captured from different devices show heterogeneous appearances because of different scan settings; thus, the DNN model trained from one domain may fail if applied directly to a new domain. As data labels are difficult to acquire, we proposed a generative adversarial network-based domain adaptation model to address the cross-domain OCT images classification task, which can extract invariant and discriminative characteristics shared by different domains without incurring additional labeling cost. A feature generator, a Wasserstein distance estimator, a domain discriminator, and a classifier were included in the model to enforce the extraction of domain invariant representations. We applied the model to OCT images as well as public digit images. Results show that the model can significantly improve the classification accuracy of cross-domain images.

Cite this Paper


BibTeX
@InProceedings{pmlr-v121-wang20a, title = {Domain adaptation model for retinopathy detection from cross-domain OCT images}, author = {Wang, Jing and Chen, Yiwei and Li, Wanyue and Kong, Wen and He, Yi and Jiang, Chuihui and Shi, Guohua}, pages = {795--810}, year = {2020}, editor = {Tal Arbel and Ismail Ben Ayed and Marleen de Bruijne and Maxime Descoteaux and Herve Lombaert and Christopher Pal}, volume = {121}, series = {Proceedings of Machine Learning Research}, address = {Montreal, QC, Canada}, month = {06--08 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v121/wang20a/wang20a.pdf}, url = {http://proceedings.mlr.press/v121/wang20a.html}, abstract = {A deep neural network (DNN) can assist in retinopathy screening by automatically classifying patients into normal and abnormal categories according to optical coherence tomography (OCT) images. Typically, OCT images captured from different devices show heterogeneous appearances because of different scan settings; thus, the DNN model trained from one domain may fail if applied directly to a new domain. As data labels are difficult to acquire, we proposed a generative adversarial network-based domain adaptation model to address the cross-domain OCT images classification task, which can extract invariant and discriminative characteristics shared by different domains without incurring additional labeling cost. A feature generator, a Wasserstein distance estimator, a domain discriminator, and a classifier were included in the model to enforce the extraction of domain invariant representations. We applied the model to OCT images as well as public digit images. Results show that the model can significantly improve the classification accuracy of cross-domain images.} }
Endnote
%0 Conference Paper %T Domain adaptation model for retinopathy detection from cross-domain OCT images %A Jing Wang %A Yiwei Chen %A Wanyue Li %A Wen Kong %A Yi He %A Chuihui Jiang %A Guohua Shi %B Proceedings of the Third Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2020 %E Tal Arbel %E Ismail Ben Ayed %E Marleen de Bruijne %E Maxime Descoteaux %E Herve Lombaert %E Christopher Pal %F pmlr-v121-wang20a %I PMLR %J Proceedings of Machine Learning Research %P 795--810 %U http://proceedings.mlr.press %V 121 %W PMLR %X A deep neural network (DNN) can assist in retinopathy screening by automatically classifying patients into normal and abnormal categories according to optical coherence tomography (OCT) images. Typically, OCT images captured from different devices show heterogeneous appearances because of different scan settings; thus, the DNN model trained from one domain may fail if applied directly to a new domain. As data labels are difficult to acquire, we proposed a generative adversarial network-based domain adaptation model to address the cross-domain OCT images classification task, which can extract invariant and discriminative characteristics shared by different domains without incurring additional labeling cost. A feature generator, a Wasserstein distance estimator, a domain discriminator, and a classifier were included in the model to enforce the extraction of domain invariant representations. We applied the model to OCT images as well as public digit images. Results show that the model can significantly improve the classification accuracy of cross-domain images.
APA
Wang, J., Chen, Y., Li, W., Kong, W., He, Y., Jiang, C. & Shi, G.. (2020). Domain adaptation model for retinopathy detection from cross-domain OCT images. Proceedings of the Third Conference on Medical Imaging with Deep Learning, in PMLR 121:795-810

Related Material