Exploring Siamese Networks in Self-Supervised Fast MRI Reconstruction

TL;DR

The proposed SiamRecon achieves the state-of-the-art reconstruction accuracy in the field of self-supervised learning on both single-coil brain MRI and multi-coil knee MRI.

Abstract

Reconstructing MR images using deep neural networks from undersampled k-space data without using fully sampled training references offers significant value in practice, which is a self-supervised regression problem calling for effective prior knowledge and supervision. The Siamese architectures are motivated by the definition "invariance" and shows promising results in unsupervised visual representative learning. Building homologous transformed images and avoiding trivial solutions are two major challenges in Siamese-based self-supervised model. In this work, we explore Siamese architecture for MRI reconstruction in a self-supervised training fashion called SiamRecon. We show the proposed approach mimics an expectation maximization algorithm. The alternative optimization provide effective supervision signal and avoid collapse. The proposed SiamRecon achieves the state-of-the-art reconstruction accuracy in the field of self-supervised learning on both single-coil brain MRI and multi-coil knee MRI.

Figures (9)

• Figure 1: The architecture of the proposed SiamRecon.
• Figure 2: The PSNR results on the single-coil brain MRI dataset.
• Figure 3: The SSIM results on the single-coil brain MRI dataset.
• Figure 4: The subjective reconstruction results of 1D line-based Cartesian undersampling patterns including 1D10$\%$, 1D20$\%$ and 1D30$\%$ from top to bottom.
• Figure 5: The subjective reconstruction results of 2D Random undersampling patterns including 2D10$\%$, 2D15$\%$ and 2D20$\%$ from top to bottom.