Image reconstruction of compressed sensing MRI using graph-based redundant wavelet transform.
TL;DR: A graph-based redundant wavelet transform is introduced to sparsely represent magnetic resonance images in iterative image reconstructions and outperforms several state-of-the-art reconstruction methods in removing artifacts and achieves fewer reconstruction errors on the tested datasets.
About: This article is published in Medical Image Analysis.The article was published on 2016-01-01. It has received 150 citations till now. The article focuses on the topics: Iterative reconstruction & Wavelet transform.
Citations
More filters
TL;DR: This paper provides a deep learning-based strategy for reconstruction of CS-MRI, and bridges a substantial gap between conventional non-learning methods working only on data from a single image, and prior knowledge from large training data sets.
Abstract: Compressed sensing magnetic resonance imaging (CS-MRI) enables fast acquisition, which is highly desirable for numerous clinical applications. This can not only reduce the scanning cost and ease patient burden, but also potentially reduce motion artefacts and the effect of contrast washout, thus yielding better image quality. Different from parallel imaging-based fast MRI, which utilizes multiple coils to simultaneously receive MR signals, CS-MRI breaks the Nyquist–Shannon sampling barrier to reconstruct MRI images with much less required raw data. This paper provides a deep learning-based strategy for reconstruction of CS-MRI, and bridges a substantial gap between conventional non-learning methods working only on data from a single image, and prior knowledge from large training data sets. In particular, a novel conditional Generative Adversarial Networks-based model (DAGAN)-based model is proposed to reconstruct CS-MRI. In our DAGAN architecture, we have designed a refinement learning method to stabilize our U-Net based generator, which provides an end-to-end network to reduce aliasing artefacts. To better preserve texture and edges in the reconstruction, we have coupled the adversarial loss with an innovative content loss. In addition, we incorporate frequency-domain information to enforce similarity in both the image and frequency domains. We have performed comprehensive comparison studies with both conventional CS-MRI reconstruction methods and newly investigated deep learning approaches. Compared with these methods, our DAGAN method provides superior reconstruction with preserved perceptual image details. Furthermore, each image is reconstructed in about 5 ms, which is suitable for real-time processing.
835 citations
Cites background from "Image reconstruction of compressed ..."
..., total variation (TV) [17]–[19], discrete cosine transforms [20]–[22] and discrete wavelet transforms [23]–[25]....
[...]
TL;DR: Two versions of a novel deep learning architecture are proposed, dubbed as ADMM-CSNet, by combining the traditional model-based CS method and data-driven deep learning method for image reconstruction from sparsely sampled measurements, which achieved favorable reconstruction accuracy in fast computational speed compared with the traditional and the other deep learning methods.
Abstract: Compressive sensing (CS) is an effective technique for reconstructing image from a small amount of sampled data. It has been widely applied in medical imaging, remote sensing, image compression, etc. In this paper, we propose two versions of a novel deep learning architecture, dubbed as ADMM-CSNet, by combining the traditional model-based CS method and data-driven deep learning method for image reconstruction from sparsely sampled measurements. We first consider a generalized CS model for image reconstruction with undetermined regularizations in undetermined transform domains, and then two efficient solvers using Alternating Direction Method of Multipliers (ADMM) algorithm for optimizing the model are proposed. We further unroll and generalize the ADMM algorithm to be two deep architectures, in which all parameters of the CS model and the ADMM algorithm are discriminatively learned by end-to-end training. For both applications of fast CS complex-valued MR imaging and CS imaging of real-valued natural images, the proposed ADMM-CSNet achieved favorable reconstruction accuracy in fast computational speed compared with the traditional and the other deep learning methods.
470 citations
TL;DR: Wang et al. as discussed by the authors proposed a projected iterative soft thresholding algorithm (pISTA) and its acceleration pFISTA for CS-MRI image reconstruction, which exploit sparsity of the magnetic resonance (MR) images under the redundant representation of tight frames.
Abstract: Compressed sensing (CS) has exhibited great potential for accelerating magnetic resonance imaging (MRI). In CS-MRI, we want to reconstruct a high-quality image from very few samples in a short time. In this paper, we propose a fast algorithm, called projected iterative soft-thresholding algorithm (pISTA), and its acceleration pFISTA for CS-MRI image reconstruction. The proposed algorithms exploit sparsity of the magnetic resonance (MR) images under the redundant representation of tight frames. We prove that pISTA and pFISTA converge to a minimizer of a convex function with a balanced tight frame sparsity formulation. The pFISTA introduces only one adjustable parameter, the step size, and we provide an explicit rule to set this parameter. Numerical experiment results demonstrate that pFISTA leads to faster convergence speeds than the state-of-art counterpart does, while achieving comparable reconstruction errors. Moreover, reconstruction errors incurred by pFISTA appear insensitive to the step size.
128 citations
TL;DR: Experimental results on simulated and real magnetic resonance spectroscopy data show that the proposed approach can successfully recover full signals from very limited samples and is robust to the estimated tensor rank.
Abstract: Signals are generally modeled as a superposition of exponential functions in spectroscopy of chemistry, biology, and medical imaging. For fast data acquisition or other inevitable reasons, however, only a small amount of samples may be acquired, and thus, how to recover the full signal becomes an active research topic, but existing approaches cannot efficiently recover $N$ -dimensional exponential signals with $N\geq 3$ . In this paper, we study the problem of recovering $N$ -dimensional (particularly $N\geq 3$ ) exponential signals from partial observations, and formulate this problem as a low-rank tensor completion problem with exponential factor vectors. The full signal is reconstructed by simultaneously exploiting the CANDECOMP/PARAFAC tensor structure and the exponential structure of the associated factor vectors. The latter is promoted by minimizing an objective function involving the nuclear norm of Hankel matrices. Experimental results on simulated and real magnetic resonance spectroscopy data show that the proposed approach can successfully recover full signals from very limited samples and is robust to the estimated tensor rank.
94 citations
TL;DR: Wang et al. as discussed by the authors proposed the first deep learning model for multi-contrast CS-MRI reconstruction, which achieved information sharing through feature sharing units, which significantly reduced the number of model parameters.
Abstract: Compressed sensing (CS) theory can accelerate multi-contrast magnetic resonance imaging (MRI) by sampling fewer measurements within each contrast. However, conventional optimization-based reconstruction models suffer several limitations, including a strict assumption of shared sparse support, time-consuming optimization, and “shallow” models with difficulties in encoding the patterns contained in massive MRI data. In this paper, we propose the first deep learning model for multi-contrast CS-MRI reconstruction. We achieve information sharing through feature sharing units, which significantly reduces the number of model parameters. The feature sharing unit combines with a data fidelity unit to comprise an inference block, which are then cascaded with dense connections, allowing for efficient information transmission across different depths of the network. Experiments on various multi-contrast MRI datasets show that the proposed model outperforms both state-of-the-art single-contrast and multi-contrast MRI methods in accuracy and efficiency. We demonstrate that improved reconstruction quality can bring benefits to subsequent medical image analysis. Furthermore, the robustness of the proposed model to misregistration shows its potential in real MRI applications.
83 citations
References
More filters
TL;DR: The proposed transform utilizes the distances between the given data points to construct tree-like structures and is explored for the recovery of labels defined on point clouds and to image denoising, showing that in both cases the results are promising.
Abstract: In this paper, we propose a new redundant wavelet transform applicable to scalar functions defined on high dimensional coordinates, weighted graphs and networks. The proposed transform utilizes the distances between the given data points to construct tree-like structures. We modify the filter-bank decomposition scheme of the redundant wavelet transform by adding in each decomposition level operators that reorder the approximation coefficients. These reordering operators are derived by organizing the tree-node features so as to shorten the path that passes through these points. We explore the use of the proposed transform for the recovery of labels defined on point clouds and to image denoising, and show that in both cases the results are promising.
46 citations
"Image reconstruction of compressed ..." refers background or methods in this paper
...The distance ,m nw between vertices is measured by , ( , )m n m n m nw w= = −b b b b , (3) where ,m nw indicates the intensity based similarity between thm and thn patches and a smaller ,m nw implies less differences (Ram et al., 2011, 2012)....
[...]
...Ram et al. proposed a simplified algorithm to attain an approximate solution (Cormen et al., 2001; Ram et al., 2011, 2012, 2013), which is listed in Algorithm 1....
[...]
...Recently, a graph is formed by viewing image patches as vertices and their differences as edges, and a shortest path on the graph is found to minimize the total difference of all image patches (Ram et al., 2011, 2012)....
[...]
...The GBRWT searches similar patches in a local region first, and expands its search to the whole region if no unvisited patches remain in local areas (Ram et al., 2011, 2012, 2013)....
[...]
TL;DR: In this article, a sparsity-promoting regularized calibration method was proposed to find a GRAPPA kernel consistent with the ACS fit equations that yields jointly sparse reconstructed coil channel images.
Abstract: The amount of calibration data needed to produce images of adequate quality can prevent auto-calibrating parallel imaging reconstruction methods like generalized autocalibrating partially parallel acquisitions (GRAPPA) from achieving a high total acceleration factor. To improve the quality of calibration when the number of auto-calibration signal (ACS) lines is restricted, we propose a sparsity-promoting regularized calibration method that finds a GRAPPA kernel consistent with the ACS fit equations that yields jointly sparse reconstructed coil channel images. Several experiments evaluate the performance of the proposed method relative to unregularized and existing regularized calibration methods for both low-quality and underdetermined fits from the ACS lines. These experiments demonstrate that the proposed method, like other regularization methods, is capable of mitigating noise amplification, and in addition, the proposed method is particularly effective at minimizing coherent aliasing artifacts caused by poor kernel calibration in real data. Using the proposed method, we can increase the total achievable acceleration while reducing degradation of the reconstructed image better than existing regularized calibration methods.
44 citations
TL;DR: This work exploits two assumed properties of dynamic MRI in order to reconstruct the images from under-sampled K-space samples and proposes a novel FOCUSS based approach to solve the optimization problem.
37 citations
"Image reconstruction of compressed ..." refers background or methods in this paper
...…et al., 2014; Zhang et al., 2012), incoherent undersampling artifacts (Greiser and von Kienlin, 2003; Tsai and Nishimura, 2000), and an effective nonlinear reconstruction algorithm (Aelterman et al., 2011; Lustig et al., 2008; Majumdar and Ward, 2011, 2012; Majumdar et al., 2013; Yue et al., 2012)....
[...]
..., 2012), incoherent undersampling artifacts (Greiser and von Kienlin, 2003; Tsai and Nishimura, 2000), and an effective nonlinear reconstruction algorithm (Aelterman et al., 2011; Lustig et al., 2008; Majumdar and Ward, 2011, 2012; Majumdar et al., 2013; Yue et al., 2012)....
[...]
TL;DR: Low‐dimensional‐structure self‐learning and thresholding reconstruction allows contrast‐enhanced whole‐heart coronary MRI with acceleration as high as 4‐fold using clinically available five‐channel phased‐array coil.
Abstract: We sought to evaluate the efficacy of prospective random undersampling and low-dimensional-structure self-learning and thresholding reconstruction for highly accelerated contrast-enhanced whole-heart coronary MRI. A prospective random undersampling scheme was implemented using phase ordering to minimize artifacts due to gradient switching and was compared to a randomly undersampled acquisition with no profile ordering. This profile-ordering technique was then used to acquire contrast-enhanced whole-heart coronary MRI in 10 healthy subjects with 4-fold acceleration. Reconstructed images and the acquired zero-filled images were compared for depicted vessel length, vessel sharpness, and subjective image quality on a scale of 1 (poor) to 4 (excellent). In a pilot study, contrast-enhanced whole-heart coronary MRI was also acquired in four patients with suspected coronary artery disease with 3-fold acceleration. The undersampled images were reconstructed using low-dimensional-structure self-learning and thresholding, which showed significant improvement over the zero-filled images in both objective and subjective measures, with an overall score of 3.6 ± 0.5. Reconstructed images in patients were all diagnostic. Low-dimensional-structure self-learning and thresholding reconstruction allows contrast-enhanced whole-heart coronary MRI with acceleration as high as 4-fold using clinically available five-channel phased-array coil.
36 citations
Additional excerpts
...Many researchers focus on accelerating MRI (Akcakaya et al., 2012; Chang and Ji, 2010; Chen et al., 2010; Jim and Zhi-Pei, 2001; Singh et al., 2011; Xiuquan et al., 2003)....
[...]
09 Jun 2011
TL;DR: This work investigates improving the image reconstruction quality through redundant translational-invariant sparsifying transforms to achieve translational invariance in compressed sensing.
Abstract: A sparse representation is an essential part of compressed sensing (CS). The discrete wavelet transform has been widely used to sparsely represent magnetic resonance images for CS applications. Artifacts usually exist in CS reconstruction when the wavelet transform is used alone. In this work, we investigate improving the image reconstruction quality through redundant translational-invariant sparsifying transforms. Cycle spinning is used with the wavelet transform and overlapping patches are used with the discrete cosine transform to achieve translational invariance. Experimental results show significant improvement in artifact reduction when contrasted with non-translational invariant transforms.
28 citations
"Image reconstruction of compressed ..." refers background in this paper
...The reason of choosing SIDWT lies in that SIDWT can mitigate blocky artifacts introduced by orthogonal wavelet transform (Baker et al., 2011)....
[...]