Joint cortical registration of geometry and function using semi-supervised learning

Jian Li, Greta Tuckute, Evelina Fedorenko, Brian L Edlow, Bruce Fischl, Adrian V Dalca
Medical Imaging with Deep Learning, PMLR 227:862-876, 2024.

Abstract

Brain surface-based image registration, an important component of brain image analysis, establishes spatial correspondence between cortical surfaces. Existing iterative and learning-based approaches focus on accurate registration of folding patterns of the cerebral cortex, and assume that geometry predicts function and thus functional areas will also be well aligned. However, structure/functional variability of anatomically corresponding areas across subjects has been widely reported. In this work, we introduce a learning-based cortical registration framework, JOSA, which jointly aligns folding patterns and functional maps while simultaneously learning an optimal atlas. We demonstrate that JOSA can substantially improve registration performance in both anatomical and functional domains over existing methods. By employing a semi-supervised training strategy, the proposed framework obviates the need for functional data during inference, enabling its use in broad neuroscientific domains where functional data may not be observed. The source code of JOSA will be released to the public at https://voxelmorph.net.

Cite this Paper

BibTeX
@InProceedings{pmlr-v227-li24b, title = {Joint cortical registration of geometry and function using semi-supervised learning}, author = {Li, Jian and Tuckute, Greta and Fedorenko, Evelina and Edlow, Brian L and Fischl, Bruce and Dalca, Adrian V}, booktitle = {Medical Imaging with Deep Learning}, pages = {862--876}, 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/li24b/li24b.pdf}, url = {https://proceedings.mlr.press/v227/li24b.html}, abstract = {Brain surface-based image registration, an important component of brain image analysis, establishes spatial correspondence between cortical surfaces. Existing iterative and learning-based approaches focus on accurate registration of folding patterns of the cerebral cortex, and assume that geometry predicts function and thus functional areas will also be well aligned. However, structure/functional variability of anatomically corresponding areas across subjects has been widely reported. In this work, we introduce a learning-based cortical registration framework, JOSA, which jointly aligns folding patterns and functional maps while simultaneously learning an optimal atlas. We demonstrate that JOSA can substantially improve registration performance in both anatomical and functional domains over existing methods. By employing a semi-supervised training strategy, the proposed framework obviates the need for functional data during inference, enabling its use in broad neuroscientific domains where functional data may not be observed. The source code of JOSA will be released to the public at https://voxelmorph.net.} }
Endnote
%0 Conference Paper %T Joint cortical registration of geometry and function using semi-supervised learning %A Jian Li %A Greta Tuckute %A Evelina Fedorenko %A Brian L Edlow %A Bruce Fischl %A Adrian V Dalca %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-li24b %I PMLR %P 862--876 %U https://proceedings.mlr.press/v227/li24b.html %V 227 %X Brain surface-based image registration, an important component of brain image analysis, establishes spatial correspondence between cortical surfaces. Existing iterative and learning-based approaches focus on accurate registration of folding patterns of the cerebral cortex, and assume that geometry predicts function and thus functional areas will also be well aligned. However, structure/functional variability of anatomically corresponding areas across subjects has been widely reported. In this work, we introduce a learning-based cortical registration framework, JOSA, which jointly aligns folding patterns and functional maps while simultaneously learning an optimal atlas. We demonstrate that JOSA can substantially improve registration performance in both anatomical and functional domains over existing methods. By employing a semi-supervised training strategy, the proposed framework obviates the need for functional data during inference, enabling its use in broad neuroscientific domains where functional data may not be observed. The source code of JOSA will be released to the public at https://voxelmorph.net.
APA
Li, J., Tuckute, G., Fedorenko, E., Edlow, B.L., Fischl, B. & Dalca, A.V.. (2024). Joint cortical registration of geometry and function using semi-supervised learning. Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 227:862-876 Available from https://proceedings.mlr.press/v227/li24b.html.

Related Material