Improved dynamic MRI reconstruction by exploiting sparsity and rank-deficiency

In this paper we address the problem of dynamic MRI reconstruction from partially sampled K-space data. Our work is motivated by previous studies in this area that proposed exploiting the spatiotemporal correlation of the dynamic MRI sequence by posing the reconstruction problem as a least squares minimization regularized by sparsity and low-rank penalties. Ideally the sparsity and low-rank penalties should be represented by the l₀-norm and the rank of a matrix; however both are NP hard penalties. The previous studies used the convex l₁-norm as a surrogate for the l₀-norm and the non-convex Schatten-q norm (0 < q ≤ 1) as a surrogate for the rank of matrix. Following past research in sparse recovery, we know that non-convex l_p-norm (0 < p ≤ 1) is a better substitute for the NP hard l₀-norm than the convex l₁-norm. Motivated by these studies, we propose improvements over the previous studies by replacing the l₁-norm sparsity penalty by the l_p-norm. Thus, we reconstruct the dynamic MRI sequence by solving a least squares minimization problem regularized by l_p-norm as the sparsity penalty and Schatten-q norm as the low-rank penalty. There are no efficient algorithms to solve the said problems. In this paper, we derive efficient algorithms to solve them. The experiments have been carried out on Dynamic Contrast Enhanced (DCE) MRI datasets. Both quantitative and qualitative analysis indicates the superiority of our proposed improvement over the existing methods.     

Magnetic resonance imaging in laboratory petrophysical core analysis

Magnetic resonance imaging (MRI) is a well-known technique in medical diagnosis and materials science. In the more specialized arena of laboratory-scale petrophysical rock core analysis, the role of MRI has undergone a substantial change in focus over the last three decades. Initially, alongside the continual drive to exploit higher magnetic field strengths in MRI applications for medicine and chemistry, the same trend was followed in core analysis. However, the spatial resolution achievable in heterogeneous porous media is inherently limited due to the magnetic susceptibility contrast between solid and fluid. As a result, imaging resolution at the length-scale of typical pore diameters is not practical and so MRI of core-plugs has often been viewed as an inappropriate use of expensive magnetic resonance facilities. Recently, there has been a paradigm shift in the use of MRI in laboratory-scale core analysis. The focus is now on acquiring data in the laboratory that are directly comparable to data obtained from magnetic resonance well-logging tools (i.e., a common physics of measurement). To maintain consistency with well-logging instrumentation, it is desirable to measure distributions of transverse (_T2$T_2$) relaxation time–the industry-standard metric in well-logging–at the laboratory-scale. These _T2$T_2$ distributions can be spatially resolved over the length of a core-plug. The use of low-field magnets in the laboratory environment is optimal for core analysis not only because the magnetic field strength is closer to that of well-logging tools, but also because the magnetic susceptibility contrast is minimized, allowing the acquisition of quantitative image voxel (or pixel) intensities that are directly scalable to liquid volume. Beyond simple determination of macroscopic rock heterogeneity, it is possible to utilize the spatial resolution for monitoring forced displacement of oil by water or chemical agents, determining capillary pressure curves, and estimating wettability. The history of MRI in petrophysics is reviewed and future directions considered, including advanced data processing techniques such as compressed sensing reconstruction and Bayesian inference analysis of under-sampled data. Although this review focuses on rock core analysis, the techniques described are applicable in a wider context to porous media in general, such as cements, soils, ceramics, and catalytic materials.     

Supervised Domain Adaptation for Automated Semantic Segmentation of the Atrial Cavity

Atrial fibrillation (AF) is the most common cardiac arrhythmia. At present, cardiac ablation is the main treatment procedure for AF. To guide and plan this procedure, it is essential for clinicians to obtain patient-specific 3D geometrical models of the atria. For this, there is an interest in automatic image segmentation algorithms, such as deep learning (DL) methods, as opposed to manual segmentation, an error-prone and time-consuming method. However, to optimize DL algorithms, many annotated examples are required, increasing acquisition costs. The aim of this work is to develop automatic and high-performance computational models for left and right atrium (LA and RA) segmentation from a few labelled MRI volumetric images with a 3D Dual U-Net algorithm. For this, a supervised domain adaptation (SDA) method is introduced to infer knowledge from late gadolinium enhanced (LGE) MRI volumetric training samples (80 LA annotated samples) to a network trained with balanced steady-state free precession (bSSFP) MR images of limited number of annotations (19 RA and LA annotated samples). The resulting knowledge-transferred model SDA outperformed the same network trained from scratch in both RA (Dice equals 0.9160) and LA (Dice equals 0.8813) segmentation tasks.     

Parallel MRI at microtesla fields

Parallel imaging techniques have been widely used in high-field magnetic resonance imaging (MRI). Multiple receiver coils have been shown to improve image quality and allow accelerated image acquisition. Magnetic resonance imaging at ultra-low fields (ULF MRI) is a new imaging approach that uses SQUID (superconducting quantum interference device) sensors to measure the spatially encoded precession of pre-polarized nuclear spin populations at microtesla-range measurement fields. In this work, parallel imaging at microtesla fields is systematically studied for the first time. A seven-channel SQUID system, designed for both ULF MRI and magnetoencephalography (MEG), is used to acquire 3D images of a human hand, as well as 2D images of a large water phantom. The imaging is performed at 46 mu T measurement field with pre-polarization at 40 mT. It is shown how the use of seven channels increases imaging field of view and improves signal-to-noise ratio for the hand images. A simple procedure for approximate correction of concomitant gradient artifacts is described. Noise propagation is analyzed experimentally, and the main source of correlated noise is identified. Accelerated imaging based on one-dimensional undersampling and 1D SENSE (sensitivity encoding) image reconstruction is studied in the case of the 2D phantom. Actual threefold imaging acceleration in comparison to single-average fully encoded Fourier imaging is demonstrated. These results show that parallel imaging methods are efficient in ULF MRI, and that imaging performance of SQUID-based instruments improves substantially as the number of channels is increased.     

钆离子:周期表中最高自旋的离子

钆离子是元素周期表中自旋值最大的离子,其轨道角动量被完全淬灭,具有十分特殊的电子结构。本文以核磁共振造影剂以及分子自旋量子比特为例介绍了钆离子在前沿科技中的应用。     

Automatic correction of in-plane bulk motion artifacts in self-navigated radial MRI

Radial MRI is typically used for scans that are sensitive to unavoidable motion. While the translational motion artifact can be easily removed from the radial trajectory data by phase correction, correction of rotational motion still remains a challenge in radial MRI. We present a novel method to refocus the image corrupted by view-to-view motion in the view-interleaved radial MRI data. In this method, the error in rotational view angles was modeled as a polynomial function of the view order and the model parameters were estimated by minimizing the self-navigator image metrics such as image entropy, gradient entropy, normalized gradient squared and mean square difference. Translational motion correction was conducted by aligning the projection profiles. Simulation studies were conducted to demonstrate the robustness of both translational and rotational motion correction methods in different noise levels. The proposed method was successfully applied to correct for motion of healthy subjects. Substantial motion correction with relative error of less than 5% was achieved by using either first- or second-order model with the image metrics. This study demonstrates the potential of the method for motion-sensitive applications.     

Simultaneous magnetic resonance imaging of diffusion anisotropy and diffusion gradient

The theory of diffusion gradient-weighted MRI (DGWI) is presented in this paper. The Bloch-Torrey equation was modified to include the effect of intravoxel spatial-location variation of water diffusion (diffusion gradient) on MRI signal, in addition to the effect of intravoxel spatial-direction variation of water diffusion (diffusion anisotropy). An analytical solution for a diffusion-encoding spin-echo pulse sequence was derived. Unlike water diffusion which attenuates the image signal intensity, this newly derived solution relates the spatial gradient of the water diffusion with the phase of the image signal. This novel MRI technique directly measures both the water diffusion and its spatial gradient, and thus offers a noninvasive imaging tool to simultaneously investigate the intravoxel inhomogeneity and anisotropy of tissue structures. In addition, as demonstrated with our preliminary data, this new method may be utilized to delineate the interfaces of tissues with different diffusion. This method is an extension of the successful diffusion tensor MRI (DTI), but requires no additional data acquisition. In addition to the measured diffusion tensor, this new method provides measurements of the spatial derivatives of the three principal diffusivities of the tensor, thereby providing additional information for improving white matter fiber tractography.     

The zonal architecture of human articular cartilage described by T2 relaxation time in the presence of Gd-DTPA2-

The depth-wise variation of T(2) relaxation time is known to reflect the collagen network architecture in cartilage, while the delayed Gadolinium Enhanced MRI of Cartilage (dGEMRIC) technique is sensitive to tissue proteoglycan (PG) concentration. As the cartilage PG content varies along the tissue depth, the depth-dependent accumulation of the contrast agent may affect the inherent T(2) of cartilage in a nonconstant manner. Therefore, T(2) and dGEMRIC are typically measured in separate MRI sessions. In the present in vitro MRI study at 9.4 T, depth-wise T(2) profiles and collagenous zone thicknesses as determined from T(2) maps in the absence and presence of Gd-DTPA(2-) (T(2) and T(2Gd), respectively) were compared in samples of intact human articular cartilage (n=65). These T(2) measures were further correlated with birefringence (BF) of polarized light microscopy (PLM) to quantify the ability of MRI to predict the properties of the collagen fibril network. The reproducibility of the T(2) measurement in the current setup was also studied. Typical tri-laminar collagen network architecture was observed both with and without Gd-DTPA(2-). The inverse of BF (1/BF) correlated significantly with both T(2) and T(2Gd) (r=0.91, slope=0.56 and r=0.90, slope=0.63), respectively. The statistically significant linear correlations between zone thicknesses as determined from T(2) and T(2Gd) were r=0.55 (slope=0.49), r=0.74 (slope=0.71) and r=0.95 (slope=0.94) for superficial, middle and deep tissue zones, respectively. Reproducibility of the T(2) measurement was worst for superficial cartilage. Consistent with PLM, T(2) and T(2Gd) measurements reveal highly similar depth-dependent information on collagen network in intact human cartilage. Thus, dGEMRIC and T(2) measurements in one MRI session are feasible for intact articular cartilage in vitro.     

Laterality index in functional MRI: methodological issues

In functional magnetic resonance imaging (fMRI), hemispheric dominance is generally indicated by a measure called the laterality index (LI). The assessment of a meaningful LI measure depends on several methodological factors that should be taken into account when interpreting LI values or comparing between subjects. Principally, these include the nature of the quantification of left and right hemispheres contributions, localisation of volumes of interest within each hemisphere, dependency on statistical threshold, thresholding LI values, choice of activation and baseline conditions and reproducibility of LI values. This review discusses such methodological factors and the different approaches that have been suggested to deal with them. Although these factors are common to a range of fMRI domains, they are discussed here in the context of fMRI of the language system.     

Quantitative correlation between (1)H MRS and dynamic contrast-enhanced MRI of human breast cancer

Proton magnetic resonance spectroscopy (¹H MRS) and dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) provide functional information, including vascular volume, vascular permeability and choline (Cho) metabolism. In this study, we applied these two imaging modalities to quantitatively characterize 36 malignant breast lesions in 32 patients and analyzed the correlation between them. Cho concentration was quantified by single-voxel ¹H MRS using water as an internal reference. The measured Cho levels ranged from 0.32 to 10.47 mmol/kg, consistent with previously reported values. In 25 mass-type lesions, the Cho concentration was significantly correlated with tumor size (r=.69, P<.0002). In addition, the Cho level was found to be significantly higher in lesions presenting as mass-type lesions compared to non-mass-type diffuse enhancements (P=.035). The enhancement kinetics from tissues covered within each MRS voxel were measured and analyzed with a two-compartmental model to obtain pharmacokinetic parameters K^trans and k_ep. A significant correlation was found between the Cho level and the pharmacokinetic parameter k_ep (r=.62, P<.0001), indicating that tissues with a high Cho level have higher wash-out rates in DCE MRI. The results suggest a correlation between Cho metabolism and angiogenesis activity, which might be explained by the association of Cho with cell replication and angiogenesis required to support tumor growth.