Group Convolutional Neural Networks for DWI Segmentation

Renfei Liu, François Lauze, Erik Bekkers, Kenny Erleben, Sune Darkner
Proceedings of the First International Workshop on Geometric Deep Learning in Medical Image Analysis, PMLR 194:96-106, 2022.

Abstract

We present a Group Convolutional Network for Segmentation of Diffusion Weighted Imaging data (DWI). The network incorporates group actions that are natural for this type of data, in the form of $SE(3)$ equivariant convolutions, i.e., roto-translation equivariant convolutions. The equivariance property provides an important inductive bias and may alleviate the need for data augmentation strategies. Instead of performing group equivariant convolutions via spectral (Fourier-based) approaches, as is common for $SE(3)$ equivariance, we implement direct and light-weight regular group convolutions. We study the effect of equivariance and weight sharing over $SE(3)$ on performances of the networks on DWI scans from the Human Connectome project. We show how that full $SE(3)$ equivariance improves segmentations, while limiting the number of learnable parameters.

Cite this Paper

BibTeX
@InProceedings{pmlr-v194-liu22a, title = {Group Convolutional Neural Networks for DWI Segmentation}, author = {Liu, Renfei and Lauze, Fran\c{c}ois and Bekkers, Erik and Erleben, Kenny and Darkner, Sune}, booktitle = {Proceedings of the First International Workshop on Geometric Deep Learning in Medical Image Analysis}, pages = {96--106}, year = {2022}, editor = {Bekkers, Erik and Wolterink, Jelmer M. and Aviles-Rivero, Angelica}, volume = {194}, series = {Proceedings of Machine Learning Research}, month = {18 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v194/liu22a/liu22a.pdf}, url = {https://proceedings.mlr.press/v194/liu22a.html}, abstract = {We present a Group Convolutional Network for Segmentation of Diffusion Weighted Imaging data (DWI). The network incorporates group actions that are natural for this type of data, in the form of $SE(3)$ equivariant convolutions, i.e., roto-translation equivariant convolutions. The equivariance property provides an important inductive bias and may alleviate the need for data augmentation strategies. Instead of performing group equivariant convolutions via spectral (Fourier-based) approaches, as is common for $SE(3)$ equivariance, we implement direct and light-weight regular group convolutions. We study the effect of equivariance and weight sharing over $SE(3)$ on performances of the networks on DWI scans from the Human Connectome project. We show how that full $SE(3)$ equivariance improves segmentations, while limiting the number of learnable parameters.} }
Endnote
%0 Conference Paper %T Group Convolutional Neural Networks for DWI Segmentation %A Renfei Liu %A François Lauze %A Erik Bekkers %A Kenny Erleben %A Sune Darkner %B Proceedings of the First International Workshop on Geometric Deep Learning in Medical Image Analysis %C Proceedings of Machine Learning Research %D 2022 %E Erik Bekkers %E Jelmer M. Wolterink %E Angelica Aviles-Rivero %F pmlr-v194-liu22a %I PMLR %P 96--106 %U https://proceedings.mlr.press/v194/liu22a.html %V 194 %X We present a Group Convolutional Network for Segmentation of Diffusion Weighted Imaging data (DWI). The network incorporates group actions that are natural for this type of data, in the form of $SE(3)$ equivariant convolutions, i.e., roto-translation equivariant convolutions. The equivariance property provides an important inductive bias and may alleviate the need for data augmentation strategies. Instead of performing group equivariant convolutions via spectral (Fourier-based) approaches, as is common for $SE(3)$ equivariance, we implement direct and light-weight regular group convolutions. We study the effect of equivariance and weight sharing over $SE(3)$ on performances of the networks on DWI scans from the Human Connectome project. We show how that full $SE(3)$ equivariance improves segmentations, while limiting the number of learnable parameters.
APA
Liu, R., Lauze, F., Bekkers, E., Erleben, K. & Darkner, S.. (2022). Group Convolutional Neural Networks for DWI Segmentation. Proceedings of the First International Workshop on Geometric Deep Learning in Medical Image Analysis, in Proceedings of Machine Learning Research 194:96-106 Available from https://proceedings.mlr.press/v194/liu22a.html.

Related Material