Imaging transformer for MRI denoising with the SNR unit training: enabling generalization across field-strengths, imaging contrasts, and anatomy

TL;DR

This paper introduces an imaging transformer model trained in SNR units with realistic noise augmentation, enabling robust MRI denoising across various field strengths, contrasts, and anatomies, surpassing CNN performance.

Contribution

The study presents a novel SNR unit training scheme and an imaging transformer architecture that generalizes MRI denoising across diverse imaging conditions.

Findings

Improved denoising performance over CNN models.

Enhanced generalization across field strengths, contrasts, and anatomy.

Effective handling of 2D, 2D+T, and 3D MRI data.

Abstract

The ability to recover MRI signal from noise is key to achieve fast acquisition, accurate quantification, and high image quality. Past work has shown convolutional neural networks can be used with abundant and paired low and high-SNR images for training. However, for applications where high-SNR data is difficult to produce at scale (e.g. with aggressive acceleration, high resolution, or low field strength), training a new denoising network using a large quantity of high-SNR images can be infeasible. In this study, we overcome this limitation by improving the generalization of denoising models, enabling application to many settings beyond what appears in the training data. Specifically, we a) develop a training scheme that uses complex MRIs reconstructed in the SNR units (i.e., the images have a fixed noise level, SNR unit training) and augments images with realistic noise based on coil g-factor, and b) develop a novel imaging transformer (imformer) to handle 2D, 2D+T, and 3D MRIs in one model architecture. Through empirical evaluation, we show this combination improves performance compared to CNN models and improves generalization, enabling a denoising model to be used across field-strengths, image contrasts, and anatomy.