3D Medical Axial Transformer: A Lightweight Transformer Model for 3D Brain Tumor Segmentation

Cheng Liu, Hisanor Kiryu
Medical Imaging with Deep Learning, PMLR 227:799-813, 2024.

Abstract

In recent years, Transformer-based models have gained attention in the field of medical image segmentation, with research exploring ways to integrate them with established architectures such as Unet. However, the high computational demands of these models have led most current approaches to focus on segmenting 2D slices of MRI or CT images, which can limit the ability of the model to learn semantic information in the depth axis and result in output with uneven edges. Additionally, the small size of medical image datasets, particularly those for brain tumor segmentation, poses a challenge for training transformer models. To address these issues, we propose 3D Medical Axial Transformer (MAT), a lightweight, end-to-end model for 3D brain tumor segmentation that employs an axial attention mechanism to reduce computational demands and self-distillation to improve performance on small datasets. Results indicate that our approach, which has fewer parameters and a simpler structure than other models, achieves superior performance and produces clearer output boundaries, making it more suitable for clinical applications.

Cite this Paper

BibTeX
@InProceedings{pmlr-v227-liu24b, title = {3D Medical Axial Transformer: A Lightweight Transformer Model for 3D Brain Tumor Segmentation}, author = {Liu, Cheng and Kiryu, Hisanor}, booktitle = {Medical Imaging with Deep Learning}, pages = {799--813}, year = {2024}, editor = {Oguz, Ipek and Noble, Jack and Li, Xiaoxiao and Styner, Martin and Baumgartner, Christian and Rusu, Mirabela and Heinmann, Tobias and Kontos, Despina and Landman, Bennett and Dawant, Benoit}, volume = {227}, series = {Proceedings of Machine Learning Research}, month = {10--12 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v227/liu24b/liu24b.pdf}, url = {https://proceedings.mlr.press/v227/liu24b.html}, abstract = {In recent years, Transformer-based models have gained attention in the field of medical image segmentation, with research exploring ways to integrate them with established architectures such as Unet. However, the high computational demands of these models have led most current approaches to focus on segmenting 2D slices of MRI or CT images, which can limit the ability of the model to learn semantic information in the depth axis and result in output with uneven edges. Additionally, the small size of medical image datasets, particularly those for brain tumor segmentation, poses a challenge for training transformer models. To address these issues, we propose 3D Medical Axial Transformer (MAT), a lightweight, end-to-end model for 3D brain tumor segmentation that employs an axial attention mechanism to reduce computational demands and self-distillation to improve performance on small datasets. Results indicate that our approach, which has fewer parameters and a simpler structure than other models, achieves superior performance and produces clearer output boundaries, making it more suitable for clinical applications.} }
Endnote
%0 Conference Paper %T 3D Medical Axial Transformer: A Lightweight Transformer Model for 3D Brain Tumor Segmentation %A Cheng Liu %A Hisanor Kiryu %B Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2024 %E Ipek Oguz %E Jack Noble %E Xiaoxiao Li %E Martin Styner %E Christian Baumgartner %E Mirabela Rusu %E Tobias Heinmann %E Despina Kontos %E Bennett Landman %E Benoit Dawant %F pmlr-v227-liu24b %I PMLR %P 799--813 %U https://proceedings.mlr.press/v227/liu24b.html %V 227 %X In recent years, Transformer-based models have gained attention in the field of medical image segmentation, with research exploring ways to integrate them with established architectures such as Unet. However, the high computational demands of these models have led most current approaches to focus on segmenting 2D slices of MRI or CT images, which can limit the ability of the model to learn semantic information in the depth axis and result in output with uneven edges. Additionally, the small size of medical image datasets, particularly those for brain tumor segmentation, poses a challenge for training transformer models. To address these issues, we propose 3D Medical Axial Transformer (MAT), a lightweight, end-to-end model for 3D brain tumor segmentation that employs an axial attention mechanism to reduce computational demands and self-distillation to improve performance on small datasets. Results indicate that our approach, which has fewer parameters and a simpler structure than other models, achieves superior performance and produces clearer output boundaries, making it more suitable for clinical applications.
APA
Liu, C. & Kiryu, H.. (2024). 3D Medical Axial Transformer: A Lightweight Transformer Model for 3D Brain Tumor Segmentation. Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 227:799-813 Available from https://proceedings.mlr.press/v227/liu24b.html.

Related Material