Undersampled magnetic resonance image (MRI) reconstruction is typically an ill-posed linear inverse task. The time and resource intensive computations require trade offs between accuracy and speed. In addition, state-of-the-art compressed sensing (CS) analytics are not cognizant of the image diagnostic quality. To address these challenges, we propose a novel CS framework that uses generative adversarial networks (GAN) to model the (low-dimensional) manifold of high-quality MR images. Leveraging a mixture of least-squares (LS) GANs and pixel-wise ℓ1/ℓ2 cost, a deep residual network with skip connections is trained as the generator that learns to remove the aliasing artifacts by projecting onto the image manifold. The LSGAN learns the texture details, while the ℓ1/ℓ2 cost suppresses high-frequency noise. A discriminator network, which is a multilayer convolutional neural network (CNN), plays the role of a perceptual cost that is then jointly trained based on high quality MR images to score the quality of retrieved images. In the operational phase, an initial aliased estimate (e.g., simply obtained by zero-filling) is propagated into the trained generator to output the desired reconstruction. This demands very low computational overhead. Extensive evaluations are performed on a large contrast-enhanced MR dataset of pediatric patients. Images rated by expert radiologists corroborate that GANCS retrieves higher quality images with improved fine texture details compared with conventional Wavelet-based and dictionary-learning based CS schemes as well as with deeplearning based schemes using pixel-wise training. In addition, it offers reconstruction times of under a few milliseconds, which is two orders of magnitude faster than current state-of-the-art CS-MRI schemes.

Purpose To develop and assess motion correction techniques for high-resolution pediatric abdominal volumetric MR images acquired free breathing with high scan efficiency. Materials and Methods First, variable-density sampling and radial-like phase-encode ordering are incorporated into the 3D Cartesian acquisition. Second, intrinsic multi-channel Butterfly navigators are used to measure respiratory motion. Lastly, these estimates are applied for both motion-weighted data-consistency in a compressed sensing and parallel imaging reconstruction, and for nonrigid motion correction using a localized autofocusing framework. With Institutional Review Board approval and informed consent/assent, studies were performed on 22 consecutive pediatric patients. Two radiologists independently scored the images for overall image quality, degree of motion artifacts, and sharpness of hepatic vessels and the diaphragm. The results are assessed using paired Wilcoxon test and weighted kappa coefficient for inter-observer agreements. Results The complete procedure yielded significantly better overall image quality (mean score of 4.7 out of 5) when compared to using no correction (mean score of 3.4, P value < 0.05) and to using motion-weighted accelerated imaging (mean score of 3.9, P value < 0.05). With an average scan time of 28 s, the proposed method resulted in comparable image quality to conventional prospective respiratory-triggered acquisitions with an average scan time of 91 s (mean score of 4.5). Conclusion With the proposed methods, diagnosable high-resolution abdominal volumetric scans can be obtained from free breathing data acquisitions.

MR scans are sensitive to motion effects due to the scan duration. To properly suppress artifacts from non-rigid body motion, complex models with elements such as translation, rotation, shear, and scaling have been incorporated into the reconstruction pipeline. However, these techniques are computationally intensive and difficult to implement for online reconstruction. On a sufficiently small spatial scale, the different types of motion can be well-approximated as simple linear translations. This formulation allows for a practical autofocusing algorithm that locally minimizes a given motion metric – more specifically, the proposed localized gradient-entropy metric. To reduce the vast search space for an optimal solution, possible motion paths are limited to the motion measured from multi-channel navigator data. The novel navigation strategy is based on the so-called “Butterfly” navigators which are modifications to the spin-warp sequence that provide intrinsic translational motion information with negligible overhead. With a 32-channel abdominal coil, sufficient number of motion measurements were found to approximate possible linear motion paths for every image voxel. The correction scheme was applied to free-breathing abdominal patient studies. In these scans, a reduction in artifacts from complex, non-rigid motion was observed.