Search

Reducing the acquisition time is important for clinical magnetic resonance imaging (MRI). Compressed sensing has recently emerged as a theoretical foundation for the reconstruction of magnetic resonance images from undersampled k-space measurements, assuming those images are sparse in a certain transform domain. However, most real-world signals are compressible rather than exactly sparse. For example, the commonly used two-dimensional wavelet for compressed sensing MRI (CS-MRI) does not sparsely represent curves and edges. In this article, we introduce a geometric image transform, the contourlet, to overcome this shortage. In addition, the improved redundancy provided by the contourlet can successfully suppress the pseudo-Gibbs phenomenon, a tiresome artefact produced by undersampling of k-space, around the singularities of images. For numerical calculation, a simple but effective iterative thresholding algorithm is employed to solve l1 norm optimization for CS-MRI. Considering the recovered information and image features, we introduce three objective criteria, which are the peak signal-to-noise ratio (PSNR), mutual information and transferred edge information, to evaluate the performance of different image transforms. Simulation results demonstrate that contourlet-based CS-MRI can better reconstruct the curves and edges than traditional wavelet-based methods, especially at low k-space sampling rate. [ABSTRACT FROM AUTHOR]

• ERRN is based on a recursive residual network and enhanced by user-designed functional modules, i.e. high-frequency feature guidance, application-specific error-correction. • The feature guidance is designed to predict the underlying anatomy based on image a priori, playing a complementary role to residual learning. • An application-specific error-correction is adapted to include different reconstruction tasks, i.e. data consistency for CS-MRI and back projection for SR-MRI. • ERRN can achieve good performance on undersampled MRI reconstruction with reduced overfitting in generalization. When using aggressive undersampling, it is difficult to recover the high quality image with reliably fine features. In this paper, we propose an enhanced recursive residual network (ERRN) that improves the basic recursive residual network with a high-frequency feature guidance, an error-correction unit and dense connections. The feature guidance is designed to predict the underlying anatomy based on image a priori learned from the label data, playing a complementary role to the residual learning. The ERRN is adapted for two important applications: compressed sensing (CS) MRI and super resolution (SR) MRI, while an application-specific error-correction unit is added into the framework, i.e. data consistency for CS-MRI and back projection for SR-MRI due to their different sampling schemes. Our proposed network was evaluated using a real-valued brain dataset, a complex-valued knee dataset, pathological brain data and in vivo rat brain data with different undersampling masks and rates. Experimental results demonstrated that ERRN presented superior reconstructions at all cases with distinctly restored structural features and highest image quality metrics compared to both the state-of-the-art convolutional neural networks and the conventional optimization-based methods, particularly for the undersampling rate over 5-fold. Thus, an excellent framework design can endow the network with a flexible architecture, fewer parameters, outstanding performances for various undersampling schemes, and reduced overfitting in generalization, which will facilitate real-time reconstruction on MRI scanners. [ABSTRACT FROM AUTHOR]

Recently, magnetic resonance imaging (MRI) reconstruction based on deep learning has become popular. Nevertheless, reconstruction of highly undersampled MR images is still challenging due to severe aliasing effects. In this study we built a deep fusion connection network (DFCN) to efficiently utilize the correlation information between adjacent slices. The proposed method was evaluated with online public IXI dataset and Calgary-Campinas-359 dataset. The results show that DFCN can generate the best reconstruction images in de-aliasing and restoring tissue structure compared with several state-of-the-art methods. The mean value of the peak signal-to-noise ratio could reach 34.16 dB, the mean value of the structural similarity is 0.9626, and the mean value of the normalized mean square error is 0.1144 on T 2 -weighted brain data of IXI dataset under 10× acceleration. Additionally, the mean value of the peak signal-to-noise ratio could reach 30.17 dB, the mean value of the structural similarity is 0.9259, and the mean value of the normalized mean square error is 0.1294 on T 1 -weighted brain data of Calgary-Campinas-359 dataset under 10× acceleration. With the correlation information between adjacent slices as prior knowledge, our method can dramatically eliminate aliasing effects and enhance the reconstruction quality of undersampled MR images. [ABSTRACT FROM AUTHOR]

Compressed sensing is utilized with the aims of reconstructing an MRI using a fraction of measurements to accelerate magnetic resonance imaging called compressed sensing magnetic resonance imaging (CS-MRI). Conventional optimization-based CS-MRI methods use random under-sampling patterns and model the MRI data in the image domain as the classic CS-MRI paradigm. Instead, we design a uniform under-sampling strategy and explore the potential of modeling the MRI data directly in the measured Fourier domain. We propose a dual-domain deep lattice network (DD-DLN) for CS-MRI with variable density uniform under-sampling. We train the networks to learn the mapping between both image and frequency domains. We observe the dual networks have complementary advantages, which motivates their combination via a lattice structure. Experiments show that the proposed DD-DLN model provides promising performance in CS-MRI under the designed variable density uniform under-sampling. [ABSTRACT FROM AUTHOR]

Compressed sensing (CS) theory assures us that we can accurately reconstruct magnetic resonance images using fewer k-space measurements than the Nyquist sampling rate requires. In traditional CS-MRI inversion methods, the fact that the energy within the Fourier measurement domain is distributed non-uniformly is often neglected during reconstruction. As a result, more densely sampled low frequency information tends to dominate penalization schemes for reconstructing MRI at the expense of high frequency details. In this paper, we propose a new framework for CS-MRI inversion in which we decompose the observed k-space data into "subspaces" via sets of filters in a lossless way, and reconstruct the images in these various spaces individually using off-the-shelf algorithms. We then fuse the results to obtain the final reconstruction. In this way, we are able to focus reconstruction on frequency information within the entire k-space more equally, preserving both high and low frequency details. We demonstrate that the proposed framework is competitive with state-of-the-art methods in CS-MRI in terms of quantitative performance, and often improves an algorithm's results qualitatively compared with its direct application to k-space. Unlabelled Image • The MRI measurements distribute non-uniformly in k-space. • A divide-and-conquer (DAC) approach for compressed sensing MRI is proposed. • The DAC approach can fully utilize the statistical characteristics of each k-space subspace. • The DAC approach often improves an undersampled MRI algorithm's results. [ABSTRACT FROM AUTHOR]

• Sampling patterns and reconstruction are jointly optimized for fast MRI. • Identical morphologic information in multi-contrast images helps to accelerate MRI. • End-to-end learning enable achievement of multiple adaptive sampling patterns. • Adaptive sampling patterns improve reconstruction quality of under-sampled images. Compressed sensing (CS) has gained increased attention in magnetic resonance imaging (MRI), leveraging its efficacy to accelerate image acquisition. Incoherence measurement and non-linear reconstruction are the most crucial guarantees of accurate restoration. However, the loose link between measurement and reconstruction hinders the further improvement of reconstruction quality, i.e., the default sampling pattern is not adaptively tailored to the downstream reconstruction method. When single-contrast reconstruction (SCR) has been upgraded to its multi-contrast reconstruction (MCR) variant, the identical morphologic information as a priori source could be integrated into the reconstruction procedure. How to measure less and reconstruct effectively by using the shareable morphologic information of various contrast images is an attractive topic. An adaptive sampling (AS) based end-to-end framework (ASSCR or ASMCR) is proposed to address this issue, which simultaneously optimizes sampling patterns and reconstruction from under-sampled data in SCR or MCR scenarios. Several deep probabilistic subsampling (DPS) modules are used in AS network to construct a sampling pattern generator. In SCR and MCR, a convolution block and a data consistency layer are iteratively applied in the reconstruction network. Specifically, the learned optimal sampling pattern output from the trained AS sub-net is used for under-sampling. Incoherence measurement for single-contrast images and the combination of sampling patterns for multi-contrast data are guided by the SCR/MCR sub-net. Experiments were conducted on two single-contrast and one multi-contrast public MRI datasets. Compared with several state-of-the-art reconstruction methods, SCR results show that a learned sampling pattern brings the quality of the reconstructed image closer to the fully-sampled reference. With the addition of different contrast images, under-sampled images with higher acceleration factors could be well recovered. The proposed method could improve the reconstruction quality of under-sampled images by using adaptive sampling patterns and learning-based reconstruction. [ABSTRACT FROM AUTHOR]

Spatiotemporally encoded (SPEN) single-shot MRI is an ultrafast MRI technique proposed recently, which utilizes quadratic rather than linear phase profile to extract the spatial information. Compared to the echo planar imaging (EPI), this technique has great advantages in resisting field inhomogeneity and chemical shift effects. Super-resolved (SR) reconstruction is adopted to compensate the inherent low resolution of SPEN images. Due to insufficient sampling rate, the SR image is challenged by aliasing artifacts and edge ghosts. The existing SR algorithms always compromise in spatial resolution to suppress these undesirable artifacts. In this paper, we proposed a novel SR algorithm termed super-resolved enhancing and edge deghosting (SEED). Different from artifacts suppression methods, our algorithm aims at exploiting the relationship between aliasing artifacts and real signal. Based on this relationship, the aliasing artifacts can be eliminated without spatial resolution loss. According to the trait of edge ghosts, finite differences and high-pass filter are employed to extract the prior knowledge of edge ghosts. By combining the prior knowledge with compressed sensing, our algorithm can efficiently reduce the edge ghosts. The robustness of SEED is demonstrated by experiments under various situations. The results indicate that the SEED can provide better spatial resolution compared to state-of-the-art SR reconstruction algorithms in SPEN MRI. Theoretical analysis and experimental results also show that the SR images reconstructed by SEED have better spatial resolution than the images obtained with conventional k -space encoding methods under similar experimental condition. [ABSTRACT FROM AUTHOR]

Highlights: [•] Random sampling was used for spatiotemporally encoded single-shot MRI. [•] Random sampling was aimed at dispersing the undersampling aliasing artifacts. [•] Compressed sensing was introduced to reconstruct spatiotemporally encoded images. [•] Singular value decomposition was used to improve the robustness of reconstruction. [•] The above hybrid scheme can reduce aliasing artifacts and improve image quality. [ABSTRACT FROM AUTHOR]