MDD-UNet: Domain Adaptation for Medical Image Segmentation with Theoretical Guarantees, a Proof of Concept

Asbjørn Munk, Mads Nielsen
Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL}), PMLR 233:174-180, 2024.

Abstract

The current state-of-the art techniques for image segmentation are often based on U-Net architectures, a U-shaped encoder-decoder networks with skip connections. Despite the powerful performance, the architecture often does not perform well when used on data which has different characteristics than the data it was trained on. Many techniques for improving performance in the presence of domain shift have been developed, however typically only have loose connections to the theory of domain adaption. In this work, we propose an unsupervised domain adaptation framework for U-Nets with theoretical guarantees based on the Margin Disparity Discrepancy called the MDD-UNet. We evaluate the proposed technique on the task of hippocampus segmentation, and find that the MDD-UNet is able to learn features which are domain-invariant with no knowledge about the labels in the target domain. The MDD-UNet improves performance over the standard U-Net on 11 out of 12 combinations of datasets. This work serves as a proof of concept by demonstrating an improvement on the U-Net in it’s standard form without modern enhancements, which opens up a new avenue of studying domain adaptation for models with very large hypothesis spaces from both methodological and practical perspectives.

Cite this Paper

BibTeX
@InProceedings{pmlr-v233-munk24a, title = {{MDD}-{UN}et: Domain Adaptation for Medical Image Segmentation with Theoretical Guarantees, a Proof of Concept}, author = {Munk, Asbj{\o}rn and Nielsen, Mads}, booktitle = {Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL})}, pages = {174--180}, year = {2024}, editor = {Lutchyn, Tetiana and Ramírez Rivera, Adín and Ricaud, Benjamin}, volume = {233}, series = {Proceedings of Machine Learning Research}, month = {09--11 Jan}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v233/munk24a/munk24a.pdf}, url = {https://proceedings.mlr.press/v233/munk24a.html}, abstract = {The current state-of-the art techniques for image segmentation are often based on U-Net architectures, a U-shaped encoder-decoder networks with skip connections. Despite the powerful performance, the architecture often does not perform well when used on data which has different characteristics than the data it was trained on. Many techniques for improving performance in the presence of domain shift have been developed, however typically only have loose connections to the theory of domain adaption. In this work, we propose an unsupervised domain adaptation framework for U-Nets with theoretical guarantees based on the Margin Disparity Discrepancy called the MDD-UNet. We evaluate the proposed technique on the task of hippocampus segmentation, and find that the MDD-UNet is able to learn features which are domain-invariant with no knowledge about the labels in the target domain. The MDD-UNet improves performance over the standard U-Net on 11 out of 12 combinations of datasets. This work serves as a proof of concept by demonstrating an improvement on the U-Net in it’s standard form without modern enhancements, which opens up a new avenue of studying domain adaptation for models with very large hypothesis spaces from both methodological and practical perspectives.} }
Endnote
%0 Conference Paper %T MDD-UNet: Domain Adaptation for Medical Image Segmentation with Theoretical Guarantees, a Proof of Concept %A Asbjørn Munk %A Mads Nielsen %B Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL}) %C Proceedings of Machine Learning Research %D 2024 %E Tetiana Lutchyn %E Adín Ramírez Rivera %E Benjamin Ricaud %F pmlr-v233-munk24a %I PMLR %P 174--180 %U https://proceedings.mlr.press/v233/munk24a.html %V 233 %X The current state-of-the art techniques for image segmentation are often based on U-Net architectures, a U-shaped encoder-decoder networks with skip connections. Despite the powerful performance, the architecture often does not perform well when used on data which has different characteristics than the data it was trained on. Many techniques for improving performance in the presence of domain shift have been developed, however typically only have loose connections to the theory of domain adaption. In this work, we propose an unsupervised domain adaptation framework for U-Nets with theoretical guarantees based on the Margin Disparity Discrepancy called the MDD-UNet. We evaluate the proposed technique on the task of hippocampus segmentation, and find that the MDD-UNet is able to learn features which are domain-invariant with no knowledge about the labels in the target domain. The MDD-UNet improves performance over the standard U-Net on 11 out of 12 combinations of datasets. This work serves as a proof of concept by demonstrating an improvement on the U-Net in it’s standard form without modern enhancements, which opens up a new avenue of studying domain adaptation for models with very large hypothesis spaces from both methodological and practical perspectives.
APA
Munk, A. & Nielsen, M.. (2024). MDD-UNet: Domain Adaptation for Medical Image Segmentation with Theoretical Guarantees, a Proof of Concept. Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL}), in Proceedings of Machine Learning Research 233:174-180 Available from https://proceedings.mlr.press/v233/munk24a.html.

Related Material