Rapid Zero-shot Image Denoising for Quantitative Imaging on a Point-Of-Care 46-mT-MRI System

Motivation: Low-field MRI holds the promise of expanding access to healthcare. The low signal-to-noise ratio (SNR) poses a significant challenge to acquiring diagnostically useful information in a reasonable scanning time. Goal(s): To overcome the challenge of low SNR in low-field MRI, achieving fast, self-supervised denoising. Approach: A rapid 4D-denoising method utilizing the Zero-Shot-Noise2Noise framework is proposed, without the need for intensive network training. Results: This method provides fast denoising in just 10-20 seconds per case and significantly boosts SNR efficiency, reducing the number of measrued TIs and TEs needed for precise, high-quality T1/T2 mapping. Impact: This study's fast 4D-denoising approach revolutionizes low-field MRI by enhancing SNR without extensive training datasets, enabling faster, more efficient imaging and broadening diagnostic accessibility in resource-limited settings.