Dual-domain self-supervised learning for accelerated non-Cartesian MRI reconstruction

• Published in: Medical image analysis Vol. 81; p. 102538
• Main Authors: Zhou, Bo, Schlemper, Jo, Dey, Neel, Mohseni Salehi, Seyed Sadegh, Sheth, Kevin, Liu, Chi, Duncan, James S., Sofka, Michal
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.10.2022
• Subjects: Accelerated MRI; Dual-domain learning; Low-field portable MRI; Non-Cartesian MRI; Self-supervised learning; Accelerated MRI; Low-field portable MRI; Self-supervised learning; 41A10; 65D05; 65D17; Non-Cartesian MRI; 41A05; Dual-domain learning
• ISSN: 1361-8415; 1361-8423
• DOI: 10.1016/j.media.2022.102538

•We propose a self-supervised learning method that enables training deep networks for non-Cartesian MRI reconstruction without access to fully sampled data.•We develop a dual-domain approach for self-supervised reconstruction in both k-space and image domains.•We demonstrate successful application on a simulated non-Cartesian MRI dataset and on real-world scenarios where fully sampled data is practically infeasible. While enabling accelerated acquisition and improved reconstruction accuracy, current deep MRI reconstruction networks are typically supervised, require fully sampled data, and are limited to Cartesian sampling patterns. These factors limit their practical adoption as fully-sampled MRI is prohibitively time-consuming to acquire clinically. Further, non-Cartesian sampling patterns are particularly desirable as they are more amenable to acceleration and show improved motion robustness. To this end, we present a fully self-supervised approach for accelerated non-Cartesian MRI reconstruction which leverages self-supervision in both k-space and image domains. In training, the undersampled data are split into disjoint k-space domain partitions. For the k-space self-supervision, we train a network to reconstruct the input undersampled data from both the disjoint partitions and from itself. For the image-level self-supervision, we enforce appearance consistency obtained from the original undersampled data and the two partitions. Experimental results on our simulated multi-coil non-Cartesian MRI dataset demonstrate that DDSS can generate high-quality reconstruction that approaches the accuracy of the fully supervised reconstruction, outperforming previous baseline methods. Finally, DDSS is shown to scale to highly challenging real-world clinical MRI reconstruction acquired on a portable low-field (0.064 T) MRI scanner with no data available for supervised training while demonstrating improved image quality as compared to traditional reconstruction, as determined by a radiologist study.