Advancing Medical Image Segmentation with Self-Supervised Learning: A 3D Student-Teacher Approach for Cardiac and Neurological Imaging

Moona Mazher, Daniel C. Alexander, Abdul Qayyum, Steven A Niederer
Proceedings of The 8th International Conference on Medical Imaging with Deep Learning, PMLR 301:1107-1126, 2026.

Abstract

We propose 3D-SegSync, a self-supervised learning (SSL) framework designed to improve segmentation accuracy for both cardiac and neurological structures. It integrates a student-teacher model with a 3D Vision-LSTM (xLSTM) backbone to capture spatial dependencies in volumetric data. The SSL phase utilizes large-scale unlabeled datasets for pretraining, followed by fine-tuning on labeled data to improve segmentation across CT and MRI scans. Experimental results demonstrate that 3D-SegSync achieves consistent performance across different anatomical structures. Additionally, its ability to generalize between CT and MRI without requiring modality-specific modifications highlights its adaptability for cardiac and neurological image segmentation. Given its strong performance, 3D-SegSync has the potential to be extended to other medical image segmentation tasks in the future. Code can be found here: https://github.com/Moona-Mazher/3D-SegSync_SSL.

Cite this Paper

BibTeX
@InProceedings{pmlr-v301-mazher26a, title = {Advancing Medical Image Segmentation with Self-Supervised Learning: A 3D Student-Teacher Approach for Cardiac and Neurological Imaging}, author = {Mazher, Moona and Alexander, Daniel C. and Qayyum, Abdul and Niederer, Steven A}, booktitle = {Proceedings of The 8th International Conference on Medical Imaging with Deep Learning}, pages = {1107--1126}, year = {2026}, editor = {Tasdizen, Tolga and Elhabian, Shireen and Summers, Ronald and Chen, Chen and Koch, Lisa and Zhuang, Yan}, volume = {301}, series = {Proceedings of Machine Learning Research}, month = {09--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v301/main/assets/mazher26a/mazher26a.pdf}, url = {https://proceedings.mlr.press/v301/mazher26a.html}, abstract = {We propose 3D-SegSync, a self-supervised learning (SSL) framework designed to improve segmentation accuracy for both cardiac and neurological structures. It integrates a student-teacher model with a 3D Vision-LSTM (xLSTM) backbone to capture spatial dependencies in volumetric data. The SSL phase utilizes large-scale unlabeled datasets for pretraining, followed by fine-tuning on labeled data to improve segmentation across CT and MRI scans. Experimental results demonstrate that 3D-SegSync achieves consistent performance across different anatomical structures. Additionally, its ability to generalize between CT and MRI without requiring modality-specific modifications highlights its adaptability for cardiac and neurological image segmentation. Given its strong performance, 3D-SegSync has the potential to be extended to other medical image segmentation tasks in the future. Code can be found here: https://github.com/Moona-Mazher/3D-SegSync_SSL.} }
Endnote
%0 Conference Paper %T Advancing Medical Image Segmentation with Self-Supervised Learning: A 3D Student-Teacher Approach for Cardiac and Neurological Imaging %A Moona Mazher %A Daniel C. Alexander %A Abdul Qayyum %A Steven A Niederer %B Proceedings of The 8th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Tolga Tasdizen %E Shireen Elhabian %E Ronald Summers %E Chen Chen %E Lisa Koch %E Yan Zhuang %F pmlr-v301-mazher26a %I PMLR %P 1107--1126 %U https://proceedings.mlr.press/v301/mazher26a.html %V 301 %X We propose 3D-SegSync, a self-supervised learning (SSL) framework designed to improve segmentation accuracy for both cardiac and neurological structures. It integrates a student-teacher model with a 3D Vision-LSTM (xLSTM) backbone to capture spatial dependencies in volumetric data. The SSL phase utilizes large-scale unlabeled datasets for pretraining, followed by fine-tuning on labeled data to improve segmentation across CT and MRI scans. Experimental results demonstrate that 3D-SegSync achieves consistent performance across different anatomical structures. Additionally, its ability to generalize between CT and MRI without requiring modality-specific modifications highlights its adaptability for cardiac and neurological image segmentation. Given its strong performance, 3D-SegSync has the potential to be extended to other medical image segmentation tasks in the future. Code can be found here: https://github.com/Moona-Mazher/3D-SegSync_SSL.
APA
Mazher, M., Alexander, D.C., Qayyum, A. & Niederer, S.A.. (2026). Advancing Medical Image Segmentation with Self-Supervised Learning: A 3D Student-Teacher Approach for Cardiac and Neurological Imaging. Proceedings of The 8th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 301:1107-1126 Available from https://proceedings.mlr.press/v301/mazher26a.html.

Related Material