Adversarial Pseudo Healthy Synthesis Needs Pathology Factorization

Tian Xia, Agisilaos Chartsias, Sotirios A. Tsaftaris
Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning, PMLR 102:512-526, 2019.

Abstract

Pseudo healthy synthesis, i.e. the creation of a subject-specific ‘healthy’ image from a pathological one, could be helpful in tasks such as anomaly detection, understanding changes induced by pathology and disease or even as data augmentation. We treat this task as a factor decomposition problem: we aim to separate what appears to be healthy and where disease is (as a map). The two factors are then recombined (by a network) to reconstruct the input disease image. We train our models in an adversarial way using either paired or unpaired settings, where we pair disease images and maps (as segmentation masks) when available. We quantitatively evaluate the quality of pseudo healthy images. We show in a series of experiments, performed in ISLES and BraTS datasets, that our method is better than conditional GAN and CycleGAN, highlighting challenges in using adversarial methods in the image translation task of pseudo healthy image generation.

Cite this Paper

BibTeX
@InProceedings{pmlr-v102-xia19a, title = {Adversarial Pseudo Healthy Synthesis Needs Pathology Factorization}, author = {Xia, Tian and Chartsias, Agisilaos and Tsaftaris, Sotirios A.}, booktitle = {Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning}, pages = {512--526}, year = {2019}, editor = {Cardoso, M. Jorge and Feragen, Aasa and Glocker, Ben and Konukoglu, Ender and Oguz, Ipek and Unal, Gozde and Vercauteren, Tom}, volume = {102}, series = {Proceedings of Machine Learning Research}, month = {08--10 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v102/xia19a/xia19a.pdf}, url = {https://proceedings.mlr.press/v102/xia19a.html}, abstract = {Pseudo healthy synthesis, i.e. the creation of a subject-specific ‘healthy’ image from a pathological one, could be helpful in tasks such as anomaly detection, understanding changes induced by pathology and disease or even as data augmentation. We treat this task as a factor decomposition problem: we aim to separate what appears to be healthy and where disease is (as a map). The two factors are then recombined (by a network) to reconstruct the input disease image. We train our models in an adversarial way using either paired or unpaired settings, where we pair disease images and maps (as segmentation masks) when available. We quantitatively evaluate the quality of pseudo healthy images. We show in a series of experiments, performed in ISLES and BraTS datasets, that our method is better than conditional GAN and CycleGAN, highlighting challenges in using adversarial methods in the image translation task of pseudo healthy image generation.} }
Endnote
%0 Conference Paper %T Adversarial Pseudo Healthy Synthesis Needs Pathology Factorization %A Tian Xia %A Agisilaos Chartsias %A Sotirios A. Tsaftaris %B Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2019 %E M. Jorge Cardoso %E Aasa Feragen %E Ben Glocker %E Ender Konukoglu %E Ipek Oguz %E Gozde Unal %E Tom Vercauteren %F pmlr-v102-xia19a %I PMLR %P 512--526 %U https://proceedings.mlr.press/v102/xia19a.html %V 102 %X Pseudo healthy synthesis, i.e. the creation of a subject-specific ‘healthy’ image from a pathological one, could be helpful in tasks such as anomaly detection, understanding changes induced by pathology and disease or even as data augmentation. We treat this task as a factor decomposition problem: we aim to separate what appears to be healthy and where disease is (as a map). The two factors are then recombined (by a network) to reconstruct the input disease image. We train our models in an adversarial way using either paired or unpaired settings, where we pair disease images and maps (as segmentation masks) when available. We quantitatively evaluate the quality of pseudo healthy images. We show in a series of experiments, performed in ISLES and BraTS datasets, that our method is better than conditional GAN and CycleGAN, highlighting challenges in using adversarial methods in the image translation task of pseudo healthy image generation.
APA
Xia, T., Chartsias, A. & Tsaftaris, S.A.. (2019). Adversarial Pseudo Healthy Synthesis Needs Pathology Factorization. Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 102:512-526 Available from https://proceedings.mlr.press/v102/xia19a.html.

Related Material