Fully Unsupervised Dynamic MRI Reconstruction via Diffeo-Temporal Equivariance

TL;DR

This paper introduces an unsupervised method for dynamic MRI reconstruction that leverages geometric spatiotemporal equivariances, outperforming existing methods and adaptable to various neural network architectures.

Contribution

It proposes a novel unsupervised framework called DDEI that exploits diffeomorphic temporal equivariance for improved dynamic MRI reconstruction.

Findings

Outperforms state-of-the-art unsupervised methods like SSDU on cardiac imaging.

Is architecture-agnostic and compatible with various neural network models.

Enables faster, higher-resolution real-time MRI imaging without fully-sampled ground truth.

Abstract

Reconstructing dynamic MRI image sequences from undersampled accelerated measurements is crucial for faster and higher spatiotemporal resolution real-time imaging of cardiac motion, free breathing motion and many other applications. Classical paradigms, such as gated cine MRI, assume periodicity, disallowing imaging of true motion. Supervised deep learning methods are fundamentally flawed as, in dynamic imaging, ground truth fully-sampled videos are impossible to truly obtain. We propose an unsupervised framework to learn to reconstruct dynamic MRI sequences from undersampled measurements alone by leveraging natural geometric spatiotemporal equivariances of MRI. Dynamic Diffeomorphic Equivariant Imaging (DDEI) significantly outperforms state-of-the-art unsupervised methods such as SSDU on highly accelerated dynamic cardiac imaging. Our method is agnostic to the underlying neural network architecture and can be used to adapt the latest models and post-processing approaches. Our code and video demos are at https://github.com/Andrewwango/ddei.