Whole-body magnetic resonance imaging at 0.05 Tesla.

Despite a half-century of advancements, global magnetic resonance imaging (MRI) accessibility remains limited and uneven, hindering its full potential in health care. Initially, MRI development focused on low fields around 0.05 Tesla, but progress halted after the introduction of the 1.5 Tesla whole-body superconducting scanner in 1983. Using a permanent 0.05 Tesla magnet and deep learning for electromagnetic interference elimination, we developed a whole-body scanner that operates using a standard wall power outlet and without radiofrequency and magnetic shielding. We demonstrated its wide-ranging applicability for imaging various anatomical structures. Furthermore, we developed three-dimensional deep learning reconstruction to boost image quality by harnessing extensive high-field MRI data. These advances pave the way for affordable deep learning-powered ultra-low-field MRI scanners, addressing unmet clinical needs in diverse health care settings worldwide.

Deep learning enabled fast 3D brain MRI at 0.055 tesla.

In recent years, there has been an intensive development of portable ultralow-field magnetic resonance imaging (MRI) for low-cost, shielding-free, and point-of-care applications. However, its quality is poor and scan time is long. We propose a fast acquisition and deep learning reconstruction framework to accelerate brain MRI at 0.055 tesla. The acquisition consists of a single average three-dimensional (3D) encoding with 2D partial Fourier sampling, reducing the scan time of T1- and T2-weighted imaging protocols to 2.5 and 3.2 minutes, respectively. The 3D deep learning leverages the homogeneous brain anatomy available in high-field human brain data to enhance image quality, reduce artifacts and noise, and improve spatial resolution to synthetic 1.5-mm isotropic resolution. Our method successfully overcomes low-signal barrier, reconstructing fine anatomical structures that are reproducible within subjects and consistent across two protocols. It enables fast and quality whole-brain MRI at 0.055 tesla, with potential for widespread biomedical applications.

Pushing the limits of low-cost ultra-low-field MRI by dual-acquisition deep learning 3D superresolution.

PURPOSE: Recent development of ultra-low-field (ULF) MRI presents opportunities for low-power, shielding-free, and portable clinical applications at a fraction of the cost. However, its performance remains limited by poor image quality. Here, a computational approach is formulated to advance ULF MR brain imaging through deep learning of large-scale publicly available 3T brain data. METHODS: A dual-acquisition 3D superresolution model is developed for ULF brain MRI at 0.055 T. It consists of deep cross-scale feature extraction, attentional fusion of two acquisitions, and reconstruction. Models for T1 -weighted and T2 -weighted imaging were trained with 3D ULF image data sets synthesized from the high-resolution 3T brain data from the Human Connectome Project. They were applied to 0.055T brain MRI with two repetitions and isotropic 3-mm acquisition resolution in healthy volunteers, young and old, as well as patients. RESULTS: The proposed approach significantly enhanced image spatial resolution and suppressed noise/artifacts. It yielded high 3D image quality at 0.055 T for the two most common neuroimaging protocols with isotropic 1.5-mm synthetic resolution and total scan time under 20 min. Fine anatomical details were restored with intrasubject reproducibility, intercontrast consistency, and confirmed by 3T MRI. CONCLUSION: The proposed dual-acquisition 3D superresolution approach advances ULF MRI for quality brain imaging through deep learning of high-field brain data. Such strategy can empower ULF MRI for low-cost brain imaging, especially in point-of-care scenarios or/and in low-income and mid-income countries.

Parallel imaging reconstruction using spatial nulling maps.

PURPOSE: To develop a robust parallel imaging reconstruction method using spatial nulling maps (SNMs). METHODS: Parallel reconstruction using null operations (PRUNO) is a k-space reconstruction method where a k-space nulling system is derived using null-subspace bases of the calibration matrix. ESPIRiT reconstruction extends the PRUNO subspace concept by exploiting the linear relationship between signal-subspace bases and spatial coil sensitivity characteristics, yielding a hybrid-domain approach. Yet it requires empirical eigenvalue thresholding to mask the coil sensitivity information and is sensitive to signal- and null-subspace division. In this study, we combine the concepts of null-subspace PRUNO and hybrid-domain ESPIRiT to provide a more robust reconstruction method that extracts null-subspace bases of calibration matrix to calculate image-domain SNMs. Multi-channel images are reconstructed by solving an image-domain nulling system formed by SNMs that contain both coil sensitivity and finite image support information, therefore, circumventing the masking-related procedure. The proposed method was evaluated with multi-channel 2D brain and knee data and compared to ESPIRiT. RESULTS: The proposed hybrid-domain method produced quality reconstruction highly comparable to ESPIRiT with optimal manual masking. It involved no masking-related manual procedure and was tolerant of the actual division of null- and signal-subspace. Spatial regularization could be also readily incorporated to reduce noise amplification as in ESPIRiT. CONCLUSION: We provide an efficient hybrid-domain reconstruction method using multi-channel SNMs that are calculated from coil calibration data. It eliminates the need for coil sensitivity masking and is relatively insensitive to subspace separation, therefore, presenting a robust parallel imaging reconstruction procedure in practice.

Calibrationless reconstruction of uniformly-undersampled multi-channel MR data with deep learning estimated ESPIRiT maps.

PURPOSE: To develop a truly calibrationless reconstruction method that derives An Eigenvalue Approach to Autocalibrating Parallel MRI (ESPIRiT) maps from uniformly-undersampled multi-channel MR data by deep learning. METHODS: ESPIRiT, one commonly used parallel imaging reconstruction technique, forms the images from undersampled MR k-space data using ESPIRiT maps that effectively represents coil sensitivity information. Accurate ESPIRiT map estimation requires quality coil sensitivity calibration or autocalibration data. We present a U-Net based deep learning model to estimate the multi-channel ESPIRiT maps directly from uniformly-undersampled multi-channel multi-slice MR data. The model is trained using fully-sampled multi-slice axial brain datasets from the same MR receiving coil system. To utilize subject-coil geometric parameters available for each dataset, the training imposes a hybrid loss on ESPIRiT maps at the original locations as well as their corresponding locations within the standard reference multi-slice axial stack. The performance of the approach was evaluated using publicly available T1-weighed brain and cardiac data. RESULTS: The proposed model robustly predicted multi-channel ESPIRiT maps from uniformly-undersampled k-space data. They were highly comparable to the reference ESPIRiT maps directly computed from 24 consecutive central k-space lines. Further, they led to excellent ESPIRiT reconstruction performance even at high acceleration, exhibiting a similar level of errors and artifacts to that by using reference ESPIRiT maps. CONCLUSION: A new deep learning approach is developed to estimate ESPIRiT maps directly from uniformly-undersampled MR data. It presents a general strategy for calibrationless parallel imaging reconstruction through learning from the coil and protocol-specific data.

Joint denoising of diffusion-weighted images via structured low-rank patch matrix approximation.

PURPOSE: To develop a joint denoising method that effectively exploits natural information redundancy in MR DWIs via low-rank patch matrix approximation. METHODS: A denoising method is introduced to jointly reduce noise in DWI dataset by exploiting nonlocal self-similarity as well as local anatomical/structural similarity within multiple 2D DWIs acquired with the same anatomical geometry but different diffusion directions. Specifically, for each small 3D reference patch sliding within 2D DWI, nonlocal but similar patches are searched by matching image contents within entire DWI dataset and then structured into a patch matrix. The resulting patch matrices are denoised by enforcing low-rankness via weighted nuclear norm minimization and finally are back-distributed to DWI space. The proposed procedure was evaluated with simulated and in vivo brain diffusion tensor imaging (DTI) datasets and then compared to existing Marchenko-Pastur principal component analysis denoising method. RESULTS: The proposed method achieved significant noise reduction while preserving structural details in all DWIs for both simulated and in vivo datasets. Quantitative evaluation of error maps demonstrated it consistently outperformed Marchenko-Pastur principal component analysis method. Further, the denoised DWIs led to substantially improved DTI parametric maps, exhibiting significantly less noise and revealing more microstructural details. CONCLUSION: The proposed method denoises DWI dataset by utilizing both nonlocal self-similarity and local structural similarity within DWI dataset. This weighted nuclear norm minimization-based low-rank patch matrix denoising approach is effective and highly applicable to various diffusion MRI applications, including DTI as a postprocessing procedure.