A multi-scale variational neural network for accelerating motion-compensated whole-heart 3D coronary MR angiography

PurposeTo enable fast reconstruction of undersampled motion-compensated whole-heart 3D coronary magnetic resonance angiography (CMRA) by learning a multi-scale variational neural network (MS-VNN) which allows the acquisition of high-quality 1.2 × 1.2 × 1.2 mm isotropic volumes in a short and predictable scan time.MethodsEighteen healthy subjects and one patient underwent free-breathing 3D CMRA acquisition with variable density spiral-like Cartesian sampling, combined with 2D image navigators for translational motion estimation/compensation. The proposed MS-VNN learns two sets of kernels and activation functions for the magnitude and phase images of the complex-valued data. For the magnitude, a multi-scale approach is applied to better capture the small calibre of the coronaries. Ten subjects were considered for training and validation. Prospectively undersampled motion-compensated data with 5-fold and 9-fold accelerations, from the remaining 9 subjects, were used to evaluate the framework. The proposed approach was compared to Wavelet-based compressed-sensing (CS), conventional VNN, and to an additional fully-sampled (FS) scan.ResultsThe average acquisition time (m:s) was 4:11 for 5-fold, 2:34 for 9-fold acceleration and 18:55 for fully-sampled. Reconstruction time with the proposed MS-VNN was ~14 s. The proposed MS-VNN achieves higher image quality than CS and VNN reconstructions, with quantitative right coronary artery sharpness (CS:43.0%, VNN:43.9%, MS-VNN:47.0%, FS:50.67%) and vessel length (CS:7.4 cm, VNN:7.7 cm, MS-VNN:8.8 cm, FS:9.1 cm) comparable to the FS scan.ConclusionThe proposed MS-VNN enables 5-fold and 9-fold undersampled CMRA acquisitions with comparable image quality that the corresponding fully-sampled scan. The proposed framework achieves extremely fast reconstruction time and does not require tuning of regularization parameters, offering easy integration into clinical workflow.     

Cavitation characteristics of flowing low and high boiling-point perfluorocarbon phase-shift nanodroplets during focused ultrasound exposures

This work investigated and compared the dynamic cavitation characteristics between low and high boiling-point phase-shift nanodroplets (NDs) under physiologically relevant flow conditions during focused ultrasound (FUS) exposures at different peak rarefactional pressures. A passive cavitation detection (PCD) system was used to monitor cavitation activity during FUS exposure at various acoustic pressure levels. Root mean square (RMS) amplitudes of broadband noise, spectrograms of the passive cavitation detection signals, and normalized inertial cavitation dose (ICD) values were calculated. Cavitation activity of low-boiling-point perfluoropentane (PFP) NDs and high boiling-point perfluorohexane (PFH) NDs flowing at in vitro mean velocities of 0–15 cm/s were compared in a 4-mm diameter wall-less vessel in a transparent tissue-mimicking phantom. In the static state, both types of phase-shift NDs exhibit a sharp rise in cavitation intensity during initial FUS exposure. Under flow conditions, cavitation activity of the PFH NDs reached the steady state less rapidly compared to PFP NDs under the lower acoustic pressure (1.35 MPa); at the higher acoustic pressure (1.65 MPa), the RMS amplitude increased more sharply during the initial FUS exposure period. In particular, the RMS-time curves of the PFP NDs shifted upward as the mean flow velocity increased from 0 to 15 cm/s; the RMS amplitude of the PFH ND solution increased from 0 to 10 cm/s and decreased at 15 cm/s. Moreover, amplitudes of the echo signal for the low boiling-point PFP NDs were higher compared to the high boiling-point PFH NDs in the lower frequency range, whereas the inverse occurred in the higher frequency range. Both PFP and PFH NDs showed increased cavitation activity in the higher frequency under the flow condition compared to the static state, especially PFH NDs. At 1.65 MPa, normalized ICD values for PFH increased from 0.93 ± 0.03 to 0.96 ± 0.04 and from 0 to 10 cm/s, then decreased to 0.86 ± 0.05 at 15 cm/s. This work contributes to our further understanding of cavitation characteristics of phase-shift NDs under physiologically relevant flow conditions during FUS exposure. In addition, the results provide a reference for selecting suitable phase-shift NDs to enhance the efficiency of cavitation-mediated ultrasonic applications.     

Rapid acquisition of magnetic resonance imaging of the shoulder using three-dimensional fast spin echo sequence with compressed sensing

PurposeTo evaluate the feasibility of 3D fast spin-echo (FSE) imaging with compressed sensing (CS) for the assessment of shoulder.Materials and methodsTwenty-nine patients who underwent shoulder MRI including image sets of axial 3D-FSE sequence without CS and with CS, using an acceleration factor of 1.5, were included. Quantitative assessment was performed by calculating the root mean square error (RMSE) and structural similarity index (SSIM). Two musculoskeletal radiologists compared image quality of 3D-FSE sequences without CS and with CS, and scored the qualitative agreement between sequences, using a five-point scale. Diagnostic agreement for pathologic shoulder lesions between the two sequences was evaluated.ResultsThe acquisition time of 3D-FSE MRI was reduced using CS (3 min 23 s vs. 2 min 22 s). Quantitative evaluations showed a significant correlation between the two sequences (r = 0.872–0.993, p < 0.05) and SSIM was in an acceptable range (0.940–0.993; mean ± standard deviation, 0.968 ± 0.018). Qualitative image quality showed good to excellent agreement between 3D-FSE images without CS and with CS. Diagnostic agreement for pathologic shoulder lesions between the two sequences was very good (κ = 0.915–1).ConclusionsThe 3D-FSE sequence with CS is feasible in evaluating the shoulder joint with reduced scan time compared to 3D-FSE without CS.     

Tagged cine magnetic resonance imaging to quantify regional mechanical changes after acute myocardial infarction

PurposeThe conventional volumetric approaches of measuring cardiac function are load-dependent, and are not able to discriminate functional changes in the infarct, transition and remote myocardium. We examined phase-dependent regional mechanical changes in the infarct, transition and remote regions after acute myocardial infarction (MI) in a preclinical mouse model using cardiovascular magnetic resonance imaging (CMR).MethodsWe induced acute MI in six mice with left anterior descending coronary artery ligation. We then examined cardiac (infarct, transition and remote-zone) morphology and function utilizing 9.4 T high field CMR before and 2 weeks after the induction of acute MI. Myocardial scar tissue was evaluated by using CMR with late gadolinium enhancement (LGE). After determining global function through volumetric analysis, regional wall motion was evaluated by measuring wall thickening and radial velocities. Strain rate imaging was performed to assess circumferential contraction and relaxation at the myocardium, endocardium, and epicardium.ResultsThere was abnormal LGE in the anterior walls after acute MI suggesting a successful MI procedure. The transition zone consisted of a mixed signal intensity, while the remote zone contained viable myocardium. As expected, the infarct zone had demonstrated severely decreased myocardial velocities and strain rates, suggesting reduced contraction and relaxation function. Compared to pre-infarct baseline, systolic and diastolic velocities (v_S and v_D) were significantly reduced at the transition zone (v_S: −1.86 ± 0.16 cm/s vs −0.68 ± 0.13 cm/s, P < 0.001; v_D: 1.86 ± 0.17 cm/s vs 0.53 ± 0.06 cm/s, P < 0.001) and remote zone (v_S: −1.86 ± 0.16 cm/s vs −0.65 ± 0.12 cm/s, P < 0.001; v_D: 1.86 ± 0.16 cm/s vs 0.51 ± 0.04 cm/s, P < 0.001). Myocardial peak systolic and diastolic strain rates (SR_S and SR_D) were significantly lower in the transition zone (SR_S: −4.2 ± 0.3 s⁻¹ vs −1.3 ± 0.2 s⁻¹, P < 0.001; SR_D: 3.9 ± 0.3 s⁻¹ vs 1.3 ± 0.2 s⁻¹, P < 0.001) and remote zone (SR_S: −3.8 ± 0.3 s⁻¹ vs −1.4 ± 0.3 s⁻¹, P < 0.001; SR_D: 3.5 ± 0.2 s⁻¹ vs 1.5 ± 0.4 s⁻¹, P = 0.006). Endocardial and epicardial SR_S and SR_D were similarly reduced in the transition and remote zones compared to baseline.ConclusionsThis study, for the first time, utilized state-of-the art high-field CMR algorithms in a preclinical mouse model for a comprehensive and controlled evaluation of the regional mechanical changes in the transition and remote zones, after acute MI. Our data demonstrate that CMR can quantitatively monitor dynamic post-MI remodeling in the transition and remote zones, thereby serving as a gold standard tool for therapeutic surveillance.     

High-performance rapid MR parameter mapping using model-based deep adversarial learning

PurposeTo develop and evaluate a deep adversarial learning-based image reconstruction approach for rapid and efficient MR parameter mapping.MethodsThe proposed method provides an image reconstruction framework by combining the end-to-end convolutional neural network (CNN) mapping, adversarial learning, and MR physical models. The CNN performs direct image-to-parameter mapping by transforming a series of undersampled images directly into MR parameter maps. Adversarial learning is used to improve image sharpness and enable better texture restoration during the image-to-parameter conversion. An additional pathway concerning the MR signal model is added between the estimated parameter maps and undersampled k-space data to ensure the data consistency during network training. The proposed framework was evaluated on T₂ mapping of the brain and the knee at an acceleration rate R = 8 and was compared with other state-of-the-art reconstruction methods. Global and regional quantitative assessments were performed to demonstrate the reconstruction performance of the proposed method.ResultsThe proposed adversarial learning approach achieved accurate T₂ mapping up to R = 8 in brain and knee joint image datasets. Compared to conventional reconstruction approaches that exploit image sparsity and low-rankness, the proposed method yielded lower errors and higher similarity to the reference and better image sharpness in the T₂ estimation. The quantitative metrics were normalized root mean square error of 3.6% for brain and 7.3% for knee, structural similarity index of 85.1% for brain and 83.2% for knee, and tenengrad measures of 9.2% for brain and 10.1% for the knee. The adversarial approach also achieved better performance for maintaining greater image texture and sharpness in comparison to the CNN approach without adversarial learning.ConclusionThe proposed framework by incorporating the efficient end-to-end CNN mapping, adversarial learning, and physical model enforced data consistency is a promising approach for rapid and efficient reconstruction of quantitative MR parameters.     

Smoothly Clipped Absolute Deviation (SCAD) regularization for compressed sensing MRI Using an augmented Lagrangian scheme

PurposeCompressed sensing (CS) provides a promising framework for MR image reconstruction from highly undersampled data, thus reducing data acquisition time. In this context, sparsity-promoting regularization techniques exploit the prior knowledge that MR images are sparse or compressible in a given transform domain. In this work, a new regularization technique was introduced by iterative linearization of the non-convex smoothly clipped absolute deviation (SCAD) norm with the aim of reducing the sampling rate even lower than it is required by the conventional l₁ norm while approaching an l₀ norm.Materials and MethodsThe CS-MR image reconstruction was formulated as an equality-constrained optimization problem using a variable splitting technique and solved using an augmented Lagrangian (AL) method developed to accelerate the optimization of constrained problems. The performance of the resulting SCAD-based algorithm was evaluated for discrete gradients and wavelet sparsifying transforms and compared with its l₁-based counterpart using phantom and clinical studies. The k-spaces of the datasets were retrospectively undersampled using different sampling trajectories. In the AL framework, the CS-MRI problem was decomposed into two simpler sub-problems, wherein the linearization of the SCAD norm resulted in an adaptively weighted soft thresholding rule with a sparsity enhancing effect.ResultsIt was demonstrated that the proposed regularization technique adaptively assigns lower weights on the thresholding of gradient fields and wavelet coefficients, and as such, is more efficient in reducing aliasing artifacts arising from k-space undersampling, when compared to its l₁-based counterpart.ConclusionThe SCAD regularization improves the performance of l₁-based regularization technique, especially at reduced sampling rates, and thus might be a good candidate for some applications in CS-MRI.     

Angiographic contrast mechanism comparison between Simultaneous Non-contrast Angiography and intraPlaque hemorrhage (SNAP) sequence and Time of Flight (TOF) sequence for intracranial artery

PurposeTo theoretically compare the MR angiography (MRA) contrast mechanism of Time of Flight (TOF) and Simultaneous Non-contrast Angiography and intraPlaque hemorrhage (SNAP) for intracranial artery imaging with in-vivo validation.MethodsThe contrast ratio (CR) of SNAP and TOF was simulated under different blood velocities and travel distance that the blood had flown through. The CR and the slope of CR with respect to blood velocity of SNAP and TOF were compared in theoretical simulation. Two healthy subjects (a 60 years old female and a 29 years old male) were imaged on a 3 T MR scanner with SNAP, TOF and phase contrast (PC) images as the validation set. The measured CR from the images in validation set was compared with the theoretically simulated CR by Person's correlation coefficient. The ratio of CR difference to velocity difference in the validation set was compared between TOF and SNAP with Student's t-test. Thirty patients (21 males, age: 48 ± 13.8 years) with carotid artery atherosclerotic plaque were imaged with both TOF and SNAP as the comparison test. Between TOF and SNAP, the CR and total artery length were compared with Student's t-test, and the prevalence of stenosis was compared with Cohen's kappa in comparison test.ResultsThe theoretically simulated CR was significantly correlated with in-vivo measured CR from the validation set for TOF (p < 0.001) and SNAP (p < 0.001). The simulation revealed that the CR of SNAP was higher than that of TOF when the blood velocity and travel distance were within the range to have effective MRA contrast. Similarly, the in-vivo comparison test showed that SNAP had higher CR (p < 0.001 for all tested intracranial arteries) and longer total artery length (1.4 ± 0.4 m vs 1.2 ± 0.2 m, p < 0.001) than TOF. The stenosis detection performance was similar between TOF and SNAP (Cohen's kappa 0.72; 95% confidence interval: 0.51–0.93). Moreover, compared with TOF, SNAP showed higher slope of CR with respect to velocity in simulation (0.06 ± 0.02 s/cm vs 0.02 ± 0.05 s/cm, p < 0.001), and higher ratio of CR difference to velocity difference in validation test (0.47 ± 0.38 s/cm vs 0.19 ± 0.38 s/cm, p = 0.001).ConclusionsCompared with TOF, the SNAP shows better performance to visualize distal intracranial artery and worse performance to visualize ICA, and is more sensitive to blood velocity.     

Deep artifact suppression for spiral real-time phase contrast cardiac magnetic resonance imaging in congenital heart disease

PurposeReal-time spiral phase contrast MR (PCMR) enables rapid free-breathing assessment of flow. Target spatial and temporal resolutions require high acceleration rates often leading to long reconstruction times. Here we propose a deep artifact suppression framework for fast and accurate flow quantification.MethodsU-Nets were trained for deep artifact suppression using 520 breath-hold gated spiral PCMR aortic datasets collected in congenital heart disease patients. Two spiral trajectories (uniform and perturbed) and two losses (Mean Absolute Error -MAE- and average structural similarity index measurement -SSIM-) were compared in synthetic data in terms of MAE, peak SNR (PSNR) and SSIM. Perturbed spiral PCMR was prospectively acquired in 20 patients. Stroke Volume (SV), peak mean velocity and edge sharpness measurements were compared to Compressed Sensing (CS) and Cartesian reference.ResultsIn synthetic data, perturbed spiral consistently outperformed uniform spiral for the different image metrics. U-Net MAE showed better MAE and PSNR while U-Net SSIM showed higher SSIM based metrics.In-vivo, there were no significant differences in SV between any of the real-time reconstructions and the reference standard Cartesian data. However, U-Net SSIM had better image sharpness and lower biases for peak velocity when compared to U-Net MAE. Reconstruction of 96 frames took ~59 s for CS and 3.9 s for U-Nets.ConclusionDeep artifact suppression of complex valued images using an SSIM based loss was successfully demonstrated in a cohort of congenital heart disease patients for fast and accurate flow quantification.     

Diagnosing Aortic Stenosis Using Phase Contrast MRI: Comparison with Echocardiography Depending on the Image Plane

This study intended to suggest a better method of measuring the precise peak velocity of the bloodstream of a patient suffering from aortic stenosis (AS) with a view to improving the reliability of the magnetic resonance imaging (MRI) scan. The study targeted 23 patients whose direction of the stenotic flow jet was different from that of the blood vessel among the patients who were diagnosed with a moderate or higher level of AS on echocardiography. The phase-contrast (PC) MRI was used for quantitatively measuring the bloodstream velocity. The examination was accomplished according to the two different image plane selection methods in the same patient. After placing the image planes perpendicular to the blood vessel and the stenotic flow jet, respectively, we obtained the two velocity encoded images to calculate the peak blood velocities to compare them with the measurement results from an echocardiography scan. According to the comparison results of the peak blood velocities, echocardiography showed a mean peak blood velocity of 4.19 ± 1.05 m/s. On the PC MRI scan, the peak blood velocity was 3.11 ± 0.77 m/s when the image plane was perpendicular to the blood vessel and 3.58 ± 0.87 m/s when the image plane was placed perpendicular to the stenotic flow jet. As a result, if PC MRI is used to measure the peak blood velocity for patients with AS, then the image plane must be placed perpendicular to the stenotic flow jet, instead of perpendicular to the blood vessel, to provide a more precise value of a low blood velocity.     

Three-dimensional black-blood T2 mapping with compressed sensing and data-driven parallel imaging in the carotid artery

PurposeTo develop a 3D black-blood T₂ mapping sequence with a combination of compressed sensing (CS) and parallel imaging (PI) for carotid wall imaging.Materials and methodsA 3D black-blood fast-spin-echo (FSE) sequence for T₂ mapping with CS and PI was developed and validated. Phantom experiments were performed to assess T₂ accuracy using a Eurospin Test Object, with different combination of CS and PI acceleration factors. A 2D multi-echo FSE sequence was used as a reference to evaluate the accuracy. The concordance correlation coefficient and Bland-Altman statistics were calculated. Twelve volunteers were scanned twice to determine the repeatability of the sequence and the intraclass correlation coefficient (ICC) was reported. Wall-lumen sharpness was calculated for different CS and PI combinations. Six patients with carotid stenosis > 50% were scanned with optimised sequence. The T₂ maps were compared with multi-contrast images.ResultsPhantom scans showed good correlation in T₂ measurement between current and reference sequence (r = 0.991). No significant difference was found between different combination of CS and PI accelerations (p = 0.999). Volunteer scans showed good repeatability of T₂ measurement (ICC: 0.93, 95% CI 0.84–0.97). The mean T₂ of the healthy wall was 48.0 ± 9.5 ms. Overall plaque T₂ values from patients were 54.9 ± 12.2 ms. Recent intraplaque haemorrhage and fibrous tissue have higher T₂ values than the mean plaque T₂ values (88.1 ± 6.8 ms and 62.7 ± 9.3 ms, respectively).ConclusionThis study demonstrates the feasibility of combining CS and PI for accelerating 3D T₂ mapping in the carotid artery, with accurate T₂ measurements and good repeatability.     

Relationship between the cardiac magnetic resonance derived extracellular volume fraction and feature tracking myocardial strain in patients with non-ischemic dilated cardiomyopathy

BackgroundFeature tracking (FT) has emerged as a promising method to quantify myocardial strain using conventional cine magnetic resonance imaging (MRI). Extracellular volume fraction (ECV) by T1 mapping enables quantification of myocardial fibrosis. To date, the correlation between FT-derived left ventricular strain and ECV has not been elucidated yet. The aim of this study was to evaluate the relationship between myocardial strain by FT and ECV by T1 mapping in patients with non-ischemic dilated cardiomyopathy (NIDCM).MethodsA total of 57 patients with NIDCM (61 ± 12 years; 46 (81%) male)) and 15 controls (62 ± 11 years; 11 (73%) male)) were studied. Using a 1.5 T magnetic resonance scanner, pre- and post- T1 mapping images of the LV wall at the mid-ventricular level were acquired to calculate the ECV by a modified Look-Locker inversion recovery (MOLLI) sequence. The radial strain (RS), circumferential strain (CS), and longitudinal strain (LS) were assessed by the FT technique. The ECV and myocardial strain were compared using a 6-segment model at the mid-ventricular level.ResultsThe ECV and myocardial strain were evaluable in all 432 segments in 72 subjects. On a patient-based analysis, NIDCM patients had a significantly higher ECV (0.30 ± 0.07 vs. 0.28 ± 0.06, p = .007) and impaired myocardial strain than the control subjects (RS, 22.7 ± 10.3 vs. 30.3 ± 18.2, p < .01; CS, −6.47 ± 1.89 vs. −9.52 ± 5.15, p < .001; LS −10.2 ± 3.78 vs. −19.8 ± 4.30, p < .001, respectively). A significant linear correlation was found between the RS and ECV (r = −0.38, p < .001) and CS and ECV, (r = 0.38, p < .001). LS and ECV also correlated (r = 0.31, p < 0.001). On a segment-based analysis, there was a significant correlation between the ECV and RS and ECV and CS (all p values < .05). The intraclass correlation coefficient was good for the strain measurement (>0.80).ConclusionsIn patients with NIDCM, significant correlation was found between myocardial strain and ECV, suggesting the FT-derived myocardial strain might be useful as a non-invasive imaging marker for the detection of myocardial fibrosis without any contrast media.     

Multispectral 3D phase-encoded turbo spin-echo for imaging near metal: Limitations and possibilities demonstrated by simulations and phantom experiments

To see improvements in the imaging performance near biomaterial implants we assessed a multispectral fully phase-encoded turbo spin-echo (ms3D-PE-TSE) sequence for artifact reduction capabilities and scan time efficiency in simulation and phantom experiments.For this purpose, ms3D-PE-TSE and ms3D-TSE sequences were implemented to obtain multispectral images (± 20 kHz) of a cobalt-chromium (CoCr) knee implant embedded in agarose. In addition, a knee implant computer model and the acquired ms3D-PE-TSE images were used to investigate the possibilities for scan time acceleration using field-of-view (FOV) reduction for off-resonance frequency bins and compressed sensing reconstructions of undersampled data. Both acceleration methods were combined to acquire a + 10 kHz frequency bin in a second experiment.The obtained ms3D-PE-TSE images showed no susceptibility related artifacts, while ms3D-TSE images suffered from hyper-intensity artifacts. The limitations of ms3D-TSE were apparent in the far off-resonance regions (±[10–20] kHz) located close to the implant. The scan time calculations showed that ms3D-PE-TSE can be applied in a clinically relevant timeframe (~ 12 min), when omitting the three central frequency bins. The feasibility of CS acceleration for ms3D-PE-TSE was demonstrated using retrospective reconstructions before combining CS and rFOV imaging to decrease the scan time for the + 10 kHz frequency bin from ~ 10.9 min to ~ 3.5 min, while also increasing the spatial resolution fourfold. The temporally resolved signal of ms3D-PE-TSE proved to be useful to decrease the intensity ripples after sum-of-squares reconstructions and increase the signal-to-noise ratio.The presented results suggest that the scan time limitations of ms3D-PE-TSE can be sufficiently addressed when focusing on signal acquisitions in the direct vicinity of metal implants. Because these regions cannot be measured with existing multispectral methods, the presented ms3D-PE-TSE method may enable the detection of inflammation or (pseudo-)tumors in locations close to the implant.     

Magnetic resonance neurography of the lumbosacral plexus at 3 Tesla – CSF-suppressed imaging with submillimeter resolution by a three-dimensional turbo spin echo sequence

PurposeTo investigate magnetic resonance neurography (MRN) of the lumbosacral plexus (LSP) with cerebrospinal fluid (CSF) suppression by using submillimeter resolution for three-dimensional (3D) turbo spin echo (TSE) imaging.Materials and methodsUsing extended phase graph (EPG) analysis, the signal response of CSF was simulated considering dephasing from coherent motion for frequency-encoding voxel sizes ranging from 0.3 to 1.3 mm and for CSF velocities ranging from 0 to 4 cm/s. In-vivo MRN included 3D TSE data with frequency encoding parallel to the feet/head axis from 15 healthy adults (mean age: 28.5 ± 3.8 years, 5 females; acquisition voxel size: 2 × 2 × 2 mm³) and 16 pediatric patients (mean age: 6.7 ± 4.1 years, 7 females; acquisition voxel size: 0.7 × 0.7 × 1.4 mm³) acquired at 3 Tesla. Five of the adults were scanned repetitively with changing acquisition voxel sizes (1 × 2 × 2 mm³, 0.7 × 2× 2 mm³, and 0.5 × 2 × 2 mm³). Measurements of the bilateral ganglion of the L5 nerve root, averaged between sides, as well as the CSF in the thecal sac were obtained for all included subjects and compared between adults and pediatric patients and between voxel sizes, using a CSF-to-nerve signal ratio (CSFNR).ResultsAccording to simulations, the CSF signal is reduced along the echo train for moving spins. Specifically, it can be reduced by over 90% compared to the maximum simulated signal for flow velocities above 2 cm/s, and could be most effectively suppressed by considering a frequency-encoding voxel size of 0.8 mm or less. For in-vivo measurements, mean CSFNR was 1.52 ± 0.22 for adults and 0.10 ± 0.03 for pediatric patients (p < .0001). Differences in CSFNR were significant between measurements using a voxel size of 2 × 2 × 2 mm³ and measurements in data with reduced voxel sizes (p ≤ .0012), with submillimeter resolution (particularly 0.5 × 2 × 2 mm³) providing highest CSF suppression.ConclusionsApplying frequency-encoding voxel sizes in submillimeter range for 3D TSE imaging with frequency encoding parallel to the feet/head axis may considerably improve MRN of LSP pathology in adults in the future because of favorable CSF suppression.     

Rapid high-resolution volumetric T1 mapping using a highly accelerated stack-of-stars Look Locker technique

PurposeTo develop a fast volumetric T₁ mapping technique.Materials and methodsA stack-of-stars (SOS) Look Locker technique based on the acquisition of undersampled radial data (>30× relative to Nyquist) and an efficient multi-slab excitation scheme is presented. A principal-component based reconstruction is used to reconstruct T₁ maps. Computer simulations were performed to determine the best choice of partitions per slab and degree of undersampling. The technique was validated in phantoms against reference T₁ values measured with a 2D Cartesian inversion-recovery spin-echo technique. The SOS Look Locker technique was tested in brain (n = 4) and prostate (n = 5). Brain T₁ mapping was carried out with and without k_z acceleration and results between the two approaches were compared. Prostate T₁ mapping was compared to standard techniques. A reproducibility study was conducted in brain and prostate. Statistical analyses were performed using linear regression and Bland Altman analysis.ResultsPhantom T₁ values showed excellent correlations between SOS Look Locker and the inversion-recovery spin-echo reference (r² = 0.9965; p < 0.0001) and between SOS Look Locker with slab-selective and non-slab selective inversion pulses (r² = 0.9999; p < 0.0001). In vivo results showed that full brain T₁ mapping (1 mm³) with k_z acceleration is achieved in 4 min 21 s. Full prostate T₁ mapping (0.9 × 0.9 × 4 mm³) is achieved in 2 min 43 s. T₁ values for brain and prostate were in agreement with literature values. A reproducibility study showed coefficients of variation in the range of 0.18–0.2% (brain) and 0.15–0.18% (prostate).ConclusionA rapid volumetric T₁ mapping technique was developed. The technique enables high-resolution T₁ mapping with adequate anatomical coverage in a clinically acceptable time.     

Spatially variant regularization of lateral displacement measurement using variance

The purpose of this work is to confirm the effectiveness of our proposed spatially variant displacement component-dependent regularization for our previously developed ultrasonic two-dimensional (2D) displacement vector measurement methods, i.e., 2D cross-spectrum phase gradient method (CSPGM), 2D autocorrelation method (AM), and 2D Doppler method (DM). Generally, the measurement accuracy of lateral displacement spatially varies and the accuracy is lower than that of axial displacement that is accurate enough. This inaccurate measurement causes an instability in a 2D shear modulus reconstruction. Thus, the spatially variant lateral displacement regularization using the lateral displacement variance will be effective in obtaining an accurate lateral strain measurement and a stable shear modulus reconstruction than a conventional spatially uniform regularization. The effectiveness is verified through agar phantom experiments. The agar phantom [60 mm (height) × 100 mm (lateral width) × 40 mm (elevational width)] that has, at a depth of 10 mm, a circular cylindrical inclusion (dia. = 10 mm) of a higher shear modulus (2.95 and 1.43 × 10⁶ N/m², i.e., relative shear modulus, 2.06) is compressed in the axial direction from the upper surface of the phantom using a commercial linear array type transducer that has a nominal frequency of 7.5-MHz. Because a contrast-to-noise ratio (CNR) expresses the detectability of the inhomogeneous region in the lateral strain image and further has almost the same sense as that of signal-to-noise ratio (SNR) for strain measurement, the obtained results show that the proposed spatially variant lateral displacement regularization yields a more accurate lateral strain measurement as well as a higher detectability in the lateral strain image (e.g., CNRs and SNRs for 2D CSPGM, 2.36 vs 2.27 and 1.74 vs 1.71, respectively). Furthermore, the spatially variant lateral displacement regularization yields a more stable and more accurate 2D shear modulus reconstruction than the uniform regularization (however, for the regularized relative shear modulus reconstructions, slightly accurate, e.g., for 2D CSPGM, 1.51 vs 1.50). These results indicate that the spatially variant displacement component-dependent regularization will enable the 2D shear modulus reconstruction to be used as practical diagnostic and monitoring tools for the effectiveness of various noninvasive therapy techniques of soft tissue diseases (e.g., breast, liver cancers). Application of the regularization to the elevational displacement will also increase the stability of a three-dimensional (3D) reconstruction.     

Compressed sensing based simultaneous black- and gray-blood carotid vessel wall MR imaging

ObjectiveIn this study, we sought to demonstrate the blood suppression performance, image quality and morphological measurements for compressed sensing (CS) based simultaneous 3D black- and gray-blood imaging sequence (CS-siBLAG) in carotid vessel wall MR imaging.Materials and methodsSeven healthy volunteers and five patients were recruited. Healthy subjects underwent five CS-siBLAG scans with 1, 2, 3, 4 and 5-fold accelerations. Signal-to-tissue ratio (STR) and contrast-to-tissue ratio (CTR) were computed as the measures of flowing signal suppression performance and the image quality for black-blood imaging of the technique. Vessel lumen area (LA) and wall area (WA) were compared between fully sampled acquisition and each accelerated acquisition. Patients underwent three CS-siBLAG scans with 1, 3 and 5-fold accelerations as well as a 3D time of flight (3D TOF) scan. Two radiologists reviewed the under-sampled black- and gray-blood image quality.ResultsSTR and CTR values obtained with 2 to 5-fold accelerations were not significantly different from those with full acquisition. LA and WA measured at 2 ×, 3 ×, 4 × and 5 × were all highly correlated to the corresponding values at 1 ×. For patients imaging, two radiologists both found that the dual-contrast images at 3 × acceleration exhibited comparable image quality to that of the fully sampled acquisition, and that the images at 5 × exhibited slightly blurred vessel wall and outer vessel wall boundaries.ConclusionBy combining the CS under-sampling pattern and reconstruction, pseudo-centric phase encoding order and dual blood contrast sequences, this technique provides spatially registered black- and gray-blood images and excellent visualization for vessel wall imaging and gray-blood imaging in a short scan time.     

Accelerating MRI fat quantification using a signal model-based dictionary to assess gastric fat volume and distribution of fat fraction

To quantify intragastric fat volume and distribution with accelerated magnetic resonance (MR) imaging using signal model-based dictionaries (DICT) in comparison to conventional parallel imaging (CG-SENSE). This study was approved by the local ethics committee and written informed consent was obtained. Seven healthy subjects were imaged after intake of a lipid emulsion and data at three different time points during the gastric emptying process was acquired in order to cover a range of fat fractions. Fully sampled and prospectively undersampled image data at a reduction factor of 4 were acquired using a multi gradient echo sequence at 1.5T. Retrospectively and prospectively undersampled data were reconstructed with DICT and CG-SENSE. Image quality of the retrospectively undersampled data was assessed relative to the fully sampled reference using the root mean square error (RMSE). In order to assess the agreement of fat volumes and intragastric fat distribution, Bland-Altman analysis and linear regression were performed on the data. The RMSE in intragastric content (ΔRMSE = 0.10 ± 0.01, P < 0.001) decreased significantly with DICT relative to CG-SENSE. CG-SENSE overestimated fat volumes (bias 2.1 ± 1.3 mL; confidence limits 5.4 and − 1.1 mL) in comparison to the prospective DICT reconstruction (bias − 0.1 ± 0.7 mL; confidence limits 1.8 and − 2.0 mL). There was a good agreement in fat distribution between the images reconstructed by retrospective DICT and the reference images (regression slope: 1.01, R² = 0.961). Accelerating gastric MRI by integrating a dictionary-based signal model allows for improved image quality and increases accuracy of fat quantification during breathholds.     

Towards accelerated quantitative sodium MRI at 7 T in the skeletal muscle: Comparison of anisotropic acquisition- and compressed sensing techniques

PurposeTo compare three anisotropic acquisition schemes and three compressed sensing (CS) approaches for accelerated tissue sodium concentration (TSC) quantification using ²³Na MRI at 7 T.Materials and methodsThree anisotropic 3D-radial acquisition sequences were evaluated using simulations, phantom- and in vivo TSC measurements: An anisotropic density-adapted 3D-radial sequence (3DPR-C), a 3D acquisition-weighted density-adapted stack-of-stars sampling scheme (SOS) and a SOS approach with golden-ratio rotation (SOS-GR). Eight healthy volunteers were examined at a 7 Tesla MRI system. TSC measurements of the calf were conducted with a nominal spatial resolution of Δx = (3.0 × 3.0 × 15.0) mm³ and a field of view of (156.0 × 156.0 × 240.0) mm³ for multiple undersampling factors (USF). Three CS reconstructions were evaluated: Total variation CS (TV-CS), 3D dictionary-learning compressed sensing (3D-DLCS) and TV-CS with a block matching prior (TV-BL-CS). Results of the simulations and measurements were compared to a simulated ground truth (GT) or a fully sampled reference measurement (FS), respectively. The deviation of the mean TSC evaluated in multiple ROI (mE_GT/FS) and the normalized root-mean-squared error (NRMSE) for simulations were evaluated for CS and NUFFT reconstructions.ResultsIn simulations, the SOS-GR yielded the lowest NRMSE and mE_GT (< 4%) with NUFFT for an acquisition time (TA) of less than 2 min. CS further improved the results. In simulations and measurements, the best TSC quantification results were obtained with 3D-DLCS and SOS-GR (lowest NRMSE, mE_GT < 2.6% in simulations, mE_GT < 10.7% for phantom measurements and mE_FS < 6% in vivo) with an USF = 4.1 (TA < 2 min). TV-CS showed no or only slight improvements to NUFFT. The results of TV-BL-CS were similar to 3D-DLCS.DiscussionThe TA for TSC measurements could be reduced to less than 2 min by using adapted sequences such as SOS-GR and CS reconstruction approaches such as 3D-DLCS or TV-BL-CS, while the quantitative accuracy stays comparable to a fully sampled NUFFT reconstruction (approx. 8 min TA). In future, the lower TA could improve clinical applicability of TSC measurements.     

Fast carotid artery MR angiography with compressed sensing based three-dimensional time-of-flight sequence

ObjectiveIn this study, we sought to investigate the feasibility of fast carotid artery MR angiography (MRA) by combining three-dimensional time-of-flight (3D TOF) with compressed sensing method (CS-3D TOF).Materials and methodsA pseudo-sequential phase encoding order was developed for CS-3D TOF to generate hyper-intense vessel and suppress background tissues in under-sampled 3D k-space. Seven healthy volunteers and one patient with carotid artery stenosis were recruited for this study. Five sequential CS-3D TOF scans were implemented at 1, 2, 3, 4 and 5-fold acceleration factors for carotid artery MRA. Blood signal-to-tissue ratio (BTR) values for fully-sampled and under-sampled acquisitions were calculated and compared in seven subjects. Blood area (BA) was measured and compared between fully sampled acquisition and each under-sampled one.ResultsThere were no significant differences between the fully-sampled dataset and each under-sampled in BTR comparisons (P > 0.05 for all comparisons). The carotid vessel BAs measured from the images of CS-3D TOF sequences with 2, 3, 4 and 5-fold acceleration scans were all highly correlated with that of the fully-sampled acquisition. The contrast between blood vessels and background tissues of the images at 2 to 5-fold acceleration is comparable to that of fully sampled images. The images at 2 × to 5 × exhibit the comparable lumen definition to the corresponding images at 1 ×.ConclusionBy combining the pseudo-sequential phase encoding order, CS reconstruction, and 3D TOF sequence, this technique provides excellent visualizations for carotid vessel and calcifications in a short scan time. It has the potential to be integrated into current multiple blood contrast imaging protocol.     

Synthetic T2-weighted images of the lumbar spine derived from an accelerated T2 mapping sequence: Comparison to conventional T2w turbo spin echo

ObjectivesTo evaluate the diagnostic usefulness of synthetic T₂-weighted images of the lumbar spine derived from ten-fold undersampled k-space data using GRAPPATINI, a combination of a model-based approach for rapid T₂ and M₀ quantification (MARTINI) extended by generalized autocalibrating partial parallel acquistion (GRAPPA).Materials and methodsOverall, 58 individuals (26 female, mean age 23.3 ± 8.1 years) were examined at 3 Tesla with sagittal and axial T₂w turbo spin echo (TSE) sequences compared to synthetic T₂₋weighted contrasts derived at identical effective echo times and spatial resolutions. Two blinded readers graded disk degeneration and evaluated the lumbar intervertebral disks for present herniation or annular tear. One reader reassessed all studies after four weeks. Weighted kappa statistics were calculated to assess inter-rater and intra-rater agreement. Also, all studies were segmented manually by one reader to compute contrast ratios (CR) and contrast-to-noise ratios (CNR) of the nucleus pulposus and the annulus fibrosus.ResultsOverall, the CR_T2w was 4.45 ± 1.80 and CR_T2synth was 4.71 ± 2.14. Both correlated (rsp = 0.768;p < 0.001) and differed (0.26 ± 1.38;p = 0.002) significantly. The CNR_T2w was 1.73 ± 0.52 and CNR_T2synth was 1.63 ± 0.50. Both correlated (rsp = 0.875;p < 0.001) and differed (−0.10 ± 0.25;p < 0.001) significantly. The inter-rater agreement was substantial to almost perfect (κ = 0.808–0.925) with the intra-rater agreement also substantial to almost perfect (κ = 0.862–0.963). The area under the curve of the receiver operating characteristics assessing disk herniation or annular tear ranged from 0.787 to 0.892.ConclusionsThis study concludes that synthetic images derived by GRAPPATINI can be used for clinical routine assessment with inter-rater and intra-rater agreements comparable to conventional T₂w TSE.