Rethinking the optimization process for self-supervised model-driven MRI reconstruction

TL;DR

This paper introduces K2Calibrate, a novel K-space adaptation strategy that enhances self-supervised MRI reconstruction by reducing noise-related errors, outperforming existing methods on the FastMRI dataset.

Contribution

The paper presents K2Calibrate, a plug-and-play K-space calibration technique that improves self-supervised model-driven MRI reconstruction by mitigating noise-induced errors.

Findings

K2Calibrate outperforms five state-of-the-art methods on FastMRI.

It effectively reduces reconstruction errors caused by dependent noise.

The method is easily integrable with various model-driven deep learning approaches.

Abstract

Recovering high-quality images from undersampled measurements is critical for accelerated MRI reconstruction. Recently, various supervised deep learning-based MRI reconstruction methods have been developed. Despite the achieved promising performances, these methods require fully sampled reference data, the acquisition of which is resource-intensive and time-consuming. Self-supervised learning has emerged as a promising solution to alleviate the reliance on fully sampled datasets. However, existing self-supervised methods suffer from reconstruction errors due to the insufficient constraint enforced on the non-sampled data points and the error accumulation happened alongside the iterative image reconstruction process for model-driven deep learning reconstrutions. To address these challenges, we propose K2Calibrate, a K-space adaptation strategy for self-supervised model-driven MR reconstruction optimization. By iteratively calibrating the learned measurements, K2Calibrate can reduce the network's reconstruction deterioration caused by statistically dependent noise. Extensive experiments have been conducted on the open-source dataset FastMRI, and K2Calibrate achieves better results than five state-of-the-art methods. The proposed K2Calibrate is plug-and-play and can be easily integrated with different model-driven deep learning reconstruction methods.