Angular Super-Resolution in Diffusion MRI with a 3D Recurrent Convolutional Autoencoder

Matthew Lyon, Paul Armitage, Mauricio A Álvarez
Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, PMLR 172:834-846, 2022.

Abstract

High resolution diffusion MRI (dMRI) data is often constrained by limited scanning time in clinical settings, thus restricting the use of downstream analysis techniques that would otherwise be available. In this work we develop a 3D recurrent convolutional neural network (RCNN) capable of super-resolving dMRI volumes in the angular (q-space) domain. Our approach formulates the task of angular super-resolution as a patch-wise regression using a 3D autoencoder conditioned on target b-vectors. Within the network we use a convolutional long short term memory (ConvLSTM) cell to model the relationship between q-space samples. We compare model performance against a baseline spherical harmonic interpolation and a 1D variant of the model architecture. We show that the 3D model has the lowest error rates across different subsampling schemes and b-values. The relative performance of the 3D RCNN is greatest in the very low angular resolution domain. Code for this project is available at github.com/m-lyon/dMRI-RCNN.

Cite this Paper

BibTeX
@InProceedings{pmlr-v172-lyon22a, title = {Angular Super-Resolution in Diffusion MRI with a 3D Recurrent Convolutional Autoencoder}, author = {Lyon, Matthew and Armitage, Paul and \'Alvarez, Mauricio A}, booktitle = {Proceedings of The 5th International Conference on Medical Imaging with Deep Learning}, pages = {834--846}, year = {2022}, editor = {Konukoglu, Ender and Menze, Bjoern and Venkataraman, Archana and Baumgartner, Christian and Dou, Qi and Albarqouni, Shadi}, volume = {172}, series = {Proceedings of Machine Learning Research}, month = {06--08 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v172/lyon22a/lyon22a.pdf}, url = {https://proceedings.mlr.press/v172/lyon22a.html}, abstract = {High resolution diffusion MRI (dMRI) data is often constrained by limited scanning time in clinical settings, thus restricting the use of downstream analysis techniques that would otherwise be available. In this work we develop a 3D recurrent convolutional neural network (RCNN) capable of super-resolving dMRI volumes in the angular (q-space) domain. Our approach formulates the task of angular super-resolution as a patch-wise regression using a 3D autoencoder conditioned on target b-vectors. Within the network we use a convolutional long short term memory (ConvLSTM) cell to model the relationship between q-space samples. We compare model performance against a baseline spherical harmonic interpolation and a 1D variant of the model architecture. We show that the 3D model has the lowest error rates across different subsampling schemes and b-values. The relative performance of the 3D RCNN is greatest in the very low angular resolution domain. Code for this project is available at github.com/m-lyon/dMRI-RCNN.} }
Endnote
%0 Conference Paper %T Angular Super-Resolution in Diffusion MRI with a 3D Recurrent Convolutional Autoencoder %A Matthew Lyon %A Paul Armitage %A Mauricio A Álvarez %B Proceedings of The 5th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2022 %E Ender Konukoglu %E Bjoern Menze %E Archana Venkataraman %E Christian Baumgartner %E Qi Dou %E Shadi Albarqouni %F pmlr-v172-lyon22a %I PMLR %P 834--846 %U https://proceedings.mlr.press/v172/lyon22a.html %V 172 %X High resolution diffusion MRI (dMRI) data is often constrained by limited scanning time in clinical settings, thus restricting the use of downstream analysis techniques that would otherwise be available. In this work we develop a 3D recurrent convolutional neural network (RCNN) capable of super-resolving dMRI volumes in the angular (q-space) domain. Our approach formulates the task of angular super-resolution as a patch-wise regression using a 3D autoencoder conditioned on target b-vectors. Within the network we use a convolutional long short term memory (ConvLSTM) cell to model the relationship between q-space samples. We compare model performance against a baseline spherical harmonic interpolation and a 1D variant of the model architecture. We show that the 3D model has the lowest error rates across different subsampling schemes and b-values. The relative performance of the 3D RCNN is greatest in the very low angular resolution domain. Code for this project is available at github.com/m-lyon/dMRI-RCNN.
APA
Lyon, M., Armitage, P. & Álvarez, M.A.. (2022). Angular Super-Resolution in Diffusion MRI with a 3D Recurrent Convolutional Autoencoder. Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 172:834-846 Available from https://proceedings.mlr.press/v172/lyon22a.html.

Related Material