Impact of B 0 $$ {\mathrm{B}}_0 $$ field imperfections correction on BOLD sensitivity in 3D-SPARKLING fMRI data

PURPOSE

Static and dynamicB 0 $$ {\mathrm{B}}_0 $$ field imperfections are detrimental to functional MRI (fMRI) applications, especially at ultra-high magnetic fields (UHF). In this work, a field camera is used to assess the benefits of retrospectively correctingB 0 $$ {\mathrm{B}}_0 $$ field perturbations on Blood Oxygen Level Dependent (BOLD) sensitivity in non-Cartesian three-dimensional (3D)-SPARKLING fMRI acquisitions.

METHODS

fMRI data were acquired at 1 mm3 $$ {}^3 $$ and for a 2.4s-TR while concurrently monitoring in real-time field perturbations using a Skope Clip-on field camera in a novel experimental setting involving a shorter TR than the required minimal TR of the field probes. Measurements of the dynamic field deviations were used along with a staticΔ B 0 $$ \Delta {\mathrm{B}}_0 $$ map to retrospectively correct static and dynamic field imperfections, respectively. In order to evaluate the impact of such a correction on fMRI volumes, a comparative study was conducted on healthy volunteers.

RESULTS

Correction ofB 0 $$ {\mathrm{B}}_0 $$ deviations improved image quality and yielded between 20% and 30% increase in median temporal signal-to-noise ratio (tSNR).Using fMRI data collected during a retinotopic mapping experiment, we demonstrated a significant increase in sensitivity to the BOLD contrast and improved accuracy of the BOLD phase maps: 44% (resp., 159%) more activated voxels were retrieved when using a significance control level based on a p-value of 0.001 without correcting for multiple comparisons (resp., 0.05 with a false discovery rate correction).

CONCLUSION

3D-SPARKLING fMRI hugely benefits from static and dynamicB 0 $$ {\mathrm{B}}_0 $$ imperfections correction. However, the proposed experimental protocol is flexible enough to be deployed on a large spectrum of encoding schemes, including arbitrary non-Cartesian readouts.

Test-Retest Reproducibility of Reduced-Field-of-View Density-Weighted CRT MRSI at 3T

Quantifying an imaging modality's ability to reproduce results is important for establishing its utility. In magnetic resonance spectroscopic imaging (MRSI), new acquisition protocols are regularly introduced which improve upon their precursors with respect to signal-to-noise ratio (SNR), total acquisition duration, and nominal voxel resolution. This study has quantified the within-subject and between-subject reproducibility of one such new protocol (reduced-field-of-view density-weighted concentric ring trajectory (rFOV-DW-CRT) MRSI) by calculating the coefficient of variance of data acquired from a test-retest experiment. The posterior cingulate cortex (PCC) and the right superior corona radiata (SCR) were selected as the regions of interest (ROIs) for grey matter (GM) and white matter (WM), respectively. CVs for between-subject and within-subject were consistently around or below 15% for Glx, tCho, and Myo-Ins, and below 5% for tNAA and tCr.

Single-shot spiral diffusion-weighted imaging at 7T using expanded encoding with compressed sensing

PURPOSE

The expanded encoding model incorporates spatially- and time-varying field perturbations for correction during reconstruction. To date, these reconstructions have used the conjugate gradient method with early stopping used as implicit regularization. However, this approach is likely suboptimal for low-SNR cases like diffusion or high-resolution MRI. Here, we investigate the extent that ℓ 1 $$ {\ell}_1 $$ -wavelet regularization, or equivalently compressed sensing (CS), combined with expanded encoding improves trade-offs between spatial resolution, readout time and SNR for single-shot spiral DWI at 7T. The reconstructions were performed using our open-source graphics processing unit-enabled reconstruction toolbox, "MatMRI," that allows inclusion of the different components of the expanded encoding model, with or without CS.

METHODS

In vivo accelerated single-shot spirals were acquired with five acceleration factors (R) (2×-6×) and three in-plane spatial resolutions (1.5, 1.3, and 1.1 mm). From the in vivo reconstructions, we estimated diffusion tensors and computed fractional anisotropy maps. Then, simulations were used to quantitatively investigate and validate the impact of CS-based regularization on image quality when compared to a known ground truth.

RESULTS

In vivo reconstructions revealed improved image quality with retainment of small features when CS was used. Simulations showed that the joint use of the expanded encoding model and CS improves accuracy of image reconstructions (reduced mean-squared error) over the range of R investigated.

CONCLUSION

The expanded encoding model and CS regularization are complementary tools for single-shot spiral diffusion MRI, which enables both higher spatial resolutions and higher R.

Model-based determination of the synchronization delay between MRI and trajectory data

PURPOSE

Real-time monitoring of dynamic magnetic fields has recently become a commercially available option for measuring MRI k-space trajectories and magnetic fields induced by eddy currents in real time. However, for accurate image reconstructions, sub-microsecond synchronization between the MRI data and field dynamics (ie, k-space trajectory plus other spatially varying fields) is required. In this work, we introduce a new model-based algorithm to automatically perform this synchronization using only the MRI data and field dynamics.

METHODS

The algorithm works by enforcing consistency among the MRI data, field dynamics, and receiver sensitivity profiles by iteratively alternating between convex optimizations for (a) the image and (b) the synchronization delay. A healthy human subject was scanned at 7 T using a transmit-receive coil with integrated field probes using both single-shot spiral and EPI, and reconstructions with various synchronization delays were compared with the result of the proposed algorithm. The accuracy of the algorithm was also investigated using simulations, in which the acquisition delays for simulated acquisitions were determined using the proposed algorithm and compared with the known ground truth.

RESULTS

In the in vivo scans, the proposed algorithm minimized artifacts related to synchronization delay for both spiral and EPI acquisitions, and the computation time required was less than 30 s. The simulations demonstrated accuracy to within tens of nanoseconds.

CONCLUSIONS

The proposed algorithm can automatically determine synchronization delays between MRI data and field dynamics measured using a field probe system.

Self-calibrated through-time spiral GRAPPA for real-time, free-breathing evaluation of left ventricular function

PURPOSE

Through-time spiral GRAPPA is a real-time imaging technique that enables ungated, free-breathing evaluation of the left ventricle. However, it requires a separate fully-sampled calibration scan to calculate GRAPPA weights. A self-calibrated through-time spiral GRAPPA method is proposed that uses a specially designed spiral trajectory with interleaved arm ordering such that consecutive undersampled frames can be merged to form calibration data, eliminating the separate fully-sampled acquisition.

THEORY AND METHODS

The proposed method considers the time needed to acquire data at all points in a GRAPPA calibration kernel when using interleaved arm ordering. Using this metric, simulations were performed to design a spiral trajectory for self-calibrated GRAPPA. Data were acquired in healthy volunteers using the proposed method and a comparison electrocardiogram-gated and breath-held cine scan. Left ventricular functional values and image quality are compared.

RESULTS

A 12-arm spiral trajectory was designed with a temporal resolution of 32.72 ms/cardiac phase with an acceleration factor of 3. Functional values calculated using the proposed method and the gold-standard method were not statistically significantly different (paired t-test, p < 0.05). Image quality ratings were lower for the proposed method, with statistically significantly different ratings (Wilcoxon signed rank test, p < 0.05) for two of five image quality aspects rated (level of artifact, blood-myocardium contrast).

CONCLUSIONS

A self-calibrated through-time spiral GRAPPA reconstruction can enable ungated, free-breathing evaluation of the left ventricle in 71 s. Functional values are equivalent to a gold-standard cine technique, although some aspects of image quality may be inferior due to the real-time nature of the data collection.

Ultra-short T2 components imaging of the whole brain using 3D dual-echo UTE MRI with rosette k-space pattern

PURPOSE

This study aimed to develop a new 3D dual-echo rosette k-space trajectory, specifically designed for UTE MRI applications. The imaging of the ultra-short transverse relaxation time (uT₂ ) of brain was acquired to test the performance of the proposed UTE sequence.

THEORY AND METHODS

The rosette trajectory was developed based on rotations of a "petal-like" pattern in the k_x -k_y plane, with oscillated extensions in the k_z -direction for 3D coverage. Five healthy volunteers underwent 10 dual-echo 3D rosette UTE scans with various TEs. Dual-exponential complex model fitting was performed on the magnitude data to separate uT₂ signals, with the output of uT₂ fraction, uT₂ value, and long-T₂ value.

RESULTS

The 3D rosette dual-echo UTE sequence showed better performance than a 3D radial UTE acquisition. More significant signal intensity decay in white matter than gray matter was observed along with the TEs. The white matter regions had higher uT₂ fraction values than gray matter (10.9% ± 1.9% vs. 5.7% ± 2.4%). The uT₂ value was approximately 0.10 ms in white matter .

CONCLUSION

The higher uT₂ fraction value in white matter compared to gray matter demonstrated the ability of the proposed sequence to capture rapidly decaying signals.

Jointly Learning Non-Cartesian k-Space Trajectories and Reconstruction Networks for 2D and 3D MR Imaging through Projection

Compressed sensing in magnetic resonance imaging essentially involves the optimization of (1) the sampling pattern in k-space under MR hardware constraints and (2) image reconstruction from undersampled k-space data. Recently, deep learning methods have allowed the community to address both problems simultaneously, especially in the non-Cartesian acquisition setting. This work aims to contribute to this field by tackling some major concerns in existing approaches. Particularly, current state-of-the-art learning methods seek hardware compliant k-space sampling trajectories by enforcing the hardware constraints through additional penalty terms in the training loss. Through ablation studies, we rather show the benefit of using a projection step to enforce these constraints and demonstrate that the resulting k-space trajectories are more flexible under a projection-based scheme, which results in superior performance in reconstructed image quality. In 2D studies, our novel method trajectories present an improved image reconstruction quality at a 20-fold acceleration factor on the fastMRI data set with SSIM scores of nearly 0.92-0.95 in our retrospective studies as compared to the corresponding Cartesian reference and also see a 3-4 dB gain in PSNR as compared to earlier state-of-the-art methods. Finally, we extend the algorithm to 3D and by comparing optimization as learning-based projection schemes, we show that data-driven joint learning-based method trajectories outperform model-based methods such as SPARKLING through a 2 dB gain in PSNR and 0.02 gain in SSIM.

Optimization of through-time radial GRAPPA with coil compression and weight sharing

PURPOSE

This work proposes principal component analysis (PCA) coil compression and weight sharing to reduce acquisition and reconstruction time of through-time radial GRAPPA.

METHODS

Through-time radial GRAPPA enables ungated free-breathing motion-resolved cardiac imaging but requires a long calibration acquisition and GRAPPA weight calculation time. PCA coil compression reduces calibration data requirements and associated acquisition time, and weight sharing reduces the number of unique GRAPPA weight sets and associated weight computation time. In vivo cardiac data reconstructed with coil compression and weight sharing are compared to a gold standard to demonstrate improvement in calibration acquisition and reconstruction performance with minimal loss of image quality.

RESULTS

Coil compression from 30 physical to 12 virtual coils (90% of signal variance) decreases requisite calibration data by 60%, reducing calibration acquisition time to 6.7 s/slice from 31.5 s/slice reported in original through-time radial GRAPPA work. Resulting images have small increase in RMS error (RMSE). Reconstruction with a weight sharing factor of 8 results in eight-fold reduction in GRAPPA weight calculation time with a comparable RMSE to reconstructions with no weight sharing. Optimized parameters for coil compression and weight sharing applied to reconstructions enables images to be collected with a temporal resolution of 66 ms/frame and spatial resolution of 2.34 × 2.34 mm while reducing calibration acquisition time from 34 to 6.7 s, weight calculation time from 200 to 3 s, and weight application time 18 to 5 s.

CONCLUSION

Coil compression and weight sharing applied to through-time radial GRAPPA enables fast free-breathing ungated cardiac cine without compromising image quality.

k-Space-based coil combination via geometric deep learning for reconstruction of non-Cartesian MRSI data

PURPOSE

State-of-the-art whole-brain MRSI with spatial-spectral encoding and multichannel acquisition generates huge amounts of data, which must be efficiently processed to stay within reasonable reconstruction times. Although coil combination significantly reduces the amount of data, currently it is performed in image space at the end of the reconstruction. This prolongs reconstruction times and increases RAM requirements. We propose an alternative k-space-based coil combination that uses geometric deep learning to combine MRSI data already in native non-Cartesian k-space.

METHODS

Twelve volunteers were scanned at a 3T MR scanner with a 20-channel head coil at 10 different positions with water-unsuppressed MRSI. At the eleventh position, water-suppressed MRSI data were acquired. Data of 7 volunteers were used to estimate sensitivity maps and form a base for simulating training data. A neural network was designed and trained to remove the effect of sensitivity profiles of the coil elements from the MRSI data. The water-suppressed MRSI data of the remaining volunteers were used to evaluate the performance of the new k-space-based coil combination relative to that of a conventional image-based alternative.

RESULTS

For both approaches, the resulting metabolic ratio maps were similar. The SNR of the k-space-based approach was comparable to the conventional approach in low SNR regions, but underperformed for high SNR. The Cramér-Rao lower bounds show the same trend. The analysis of the FWHM showed no difference between the two methods.

CONCLUSION

k-Space-based coil combination of MRSI data is feasible and reduces the amount of raw data immediately after their sampling.

Quantification of lung water density with UTE Yarnball MRI

PURPOSE

An efficient Yarnball ultrashort-TE k-space trajectory, in combination with an optimized pulse sequence design and automated image-processing approach, is proposed for fast and quantitative imaging of water density in the lung parenchyma.

METHODS

Three-dimensional Yarnball k-space trajectories (TE = 0.07 ms) were designed at 3 T for breath-hold and free-breathing navigator acquisitions targeting the lung parenchyma (full torso spatial coverage) with minimal T₁ and T 2 ∗ weighting. A composite of all solid tissues surrounding the lungs (muscle, liver, heart, blood pool) was used for user-independent lung water density signal referencing and B₁ -inhomogeneity correction needed for the calculation of relative lung water density images. Sponge phantom experiments were used to validate absolute water density quantification, and relative lung water density was evaluated in 10 healthy volunteers.

RESULTS

Phantom experiments showed excellent agreement between sponge wet weight and imaging-derived water density. Breath-hold (13 seconds) and free-breathing (~2 minutes) Yarnball acquisitions in volunteers (2.5-mm isotropic resolution) had negligible artifacts and good lung parenchyma SNR (>10). Whole-lung average relative lung water density values with fully automated analysis were 28.2 ± 1.9% and 28.6 ± 1.8% for breath-hold and free-breathing acquisitions, respectively, with good test-retest reproducibility (intraclass correlation coefficient = 0.86 and 0.95, respectively).

CONCLUSIONS

Quantitative lung water density imaging with an optimized Yarnball k-space acquisition approach is possible in a breath-hold or short free-breathing study with automated signal referencing and segmentation.

Non-Cartesian k-space trajectory calculation based on concurrent reading of the gradient amplifiers' output currents

PURPOSE

Non-Cartesian imaging sequences involve sampling during rapid variation of the encoding field gradients. The quality of the reconstructed images often suffers from insufficient knowledge of the exact dynamics of the actual fields applied during sampling.

METHODS

We propose determination of the accurate field dynamics by measuring the currents at the gradient amplifier outputs using the amplifiers' internal sensors concurrently with imaging. The actual dynamic field evolution is then determined by convolution with the measured current-to-field impulse response function of the gradient coil. Integration of the gradient field evolution allows derivation of the k-space trajectory for reconstruction.

RESULTS

The current-based approach is investigated in spiral and ultrashort TE phantom imaging. In comparison with the model-based product reconstruction as well as a correction approach based on the conventional input waveform-to-field impulse response function, it provides slightly improved image quality. The improvement is ascribed to a better representation of eddy current and amplifier nonlinearity effects.

CONCLUSION

Trajectory calculation based on measured amplifier output currents offers a robust, purely measurement-based alternative to conventional model-based approaches. The implementation can mitigate gradient amplifier imperfections with no or little additional hardware effort.

Three-dimensional Yarnball k-space acquisition for accelerated MRI

PURPOSE

To introduce an efficient sampling technique named Yarnball, which may serve as a direct alternative to 3D Cones.

METHODS

Yarnball evolves through 3D k-space with increasing loop size, and the differential equations defining this flexible trajectory are presented in detail. The sampling efficiencies of Yarnball and 3D Cones were compared through point spread function analysis and simulated imaging (which highlights undersampling in the absence of other scanning effects). The feasibility of Yarnball implementation was demonstrated for fully sampled T₁ -weighted images of the human head at 3 T.

RESULTS

The mostly large 3D loops of the Yarnball trajectory facilitate rapid sampling under peripheral nerve stimulation constraint, an advantage that increases with readout duration (T_RO ). Point spread function analysis yielded 89% (T_RO = 2 ms) and 77% (T_RO = 10 ms) of Yarnball voxels with magnitude less than 0.01% of the point spread function peak. For 3D Cones, these values were only 52% and 29%. The 3D-Cones technique required 1.4 times (T_RO = 2 ms) and 1.8 times (T_RO = 10 ms) more trajectories than Yarnball to produce simulated images of a sphere free from undersampling artifact. For a prolate spheroidal (head-like) object, 1.75 times and 2.6 times more trajectories were required for 3D Cones. Yarnball produced 0.72 mm (1/2k_max ) isotropic T₁ -weighted human brain images free from undersampling artifact in only 98 seconds at 3 T.

CONCLUSION

Yarnball demonstrated greater k-space sampling efficiency than directly comparable 3D Cones, and may have value wherever 3D Cones has been considered. Yarnball may also have value in the context of rapid T₁ -weighted brain imaging.

3D variable-density SPARKLING trajectories for high-resolution T2*-weighted magnetic resonance imaging

We have recently proposed a new optimization algorithm called SPARKLING (Spreading Projection Algorithm for Rapid K-space sampLING) to design efficient compressive sampling patterns for magnetic resonance imaging (MRI). This method has a few advantages over conventional non-Cartesian trajectories such as radial lines or spirals: i) it allows to sample the k-space along any arbitrary density while the other two are restricted to radial densities and ii) it optimizes the gradient waveforms for a given readout time. Here, we introduce an extension of the SPARKLING method for 3D imaging by considering both stacks-of-SPARKLING and fully 3D SPARKLING trajectories. Our method allowed to achieve an isotropic resolution of 600 μm in just 45 seconds for T2∗-weighted ex vivo brain imaging at 7 Tesla over a field-of-view of 200 × 200 × 140 mm³ . Preliminary in vivo human brain data shows that a stack-of-SPARKLING is less subject to off-resonance artifacts than a stack-of-spirals.

Rosette Trajectories Enable Ungated, Motion-Robust, Simultaneous Cardiac and Liver T2 * Iron Assessment

BACKGROUND

Quantitative T₂ * MRI is the standard of care for the assessment of iron overload. However, patient motion corrupts T₂ * estimates.

PURPOSE

To develop and evaluate a motion-robust, simultaneous cardiac and liver T₂ * imaging approach using non-Cartesian, rosette sampling and a model-based reconstruction as compared to clinical-standard Cartesian MRI.

STUDY TYPE

Prospective.

PHANTOM/POPULATION

Six ferumoxytol-containing phantoms (26-288 μg/mL). Eight healthy subjects and 18 patients referred for clinically indicated iron overload assessment.

FIELD STRENGTH/SEQUENCE

1.5T, 2D Cartesian and rosette gradient echo (GRE) ASSESSMENT: GRE T₂ * values were validated in ferumoxytol phantoms. In healthy subjects, test-retest and spatial coefficient of variation (CoV) analysis was performed during three breathing conditions. Cartesian and rosette T₂ * were compared using correlation and Bland-Altman analysis. Images were rated by three experienced radiologists on a 5-point scale.

STATISTICAL TESTS

Linear regression, analysis of variance (ANOVA), and paired Student's t-testing were used to compare reproducibility and variability metrics in Cartesian and rosette scans. The Wilcoxon rank test was used to assess reader score comparisons and reader reliability was measured using intraclass correlation analysis.

RESULTS

Rosette R2* (1/T₂ *) was linearly correlated with ferumoxytol concentration (r² = 1.00) and not significantly different than Cartesian values (P = 0.16). During breath-holding, ungated rosette liver and heart T₂ * had lower spatial CoV (liver: 18.4 ± 9.3% Cartesian, 8.8% ± 3.4% rosette, P = 0.02, heart: 37.7% ± 14.3% Cartesian, 13.4% ± 1.7% rosette, P = 0.001) and higher-quality scores (liver: 3.3 [3.0-3.6] Cartesian, 4.7 [4.1-4.9] rosette, P = 0.005, heart: 3.0 [2.3-3] Cartesian, 4.5 [3.8-5.0] rosette, P = 0.005) compared to Cartesian values. During free-breathing and failed breath-holding, Cartesian images had very poor to average image quality with significant artifacts, whereas rosette remained very good, with minimal artifacts (P = 0.001).

DATA CONCLUSION

Rosette k-sampling with a model-based reconstruction offers a clinically useful motion-robust T₂ * mapping approach for iron quantification. J. MAGN. RESON. IMAGING 2020;52:1688-1698.

Reconstruction of undersampled 3D non-Cartesian image-based navigators for coronary MRA using an unrolled deep learning model

PURPOSE

To rapidly reconstruct undersampled 3D non-Cartesian image-based navigators (iNAVs) using an unrolled deep learning (DL) model, enabling nonrigid motion correction in coronary magnetic resonance angiography (CMRA).

METHODS

An end-to-end unrolled network is trained to reconstruct beat-to-beat 3D iNAVs acquired during a CMRA sequence. The unrolled model incorporates a nonuniform FFT operator in TensorFlow to perform the data-consistency operation, and the regularization term is learned by a convolutional neural network (CNN) based on the proximal gradient descent algorithm. The training set includes 6,000 3D iNAVs acquired from 7 different subjects and 11 scans using a variable-density (VD) cones trajectory. For testing, 3D iNAVs from 4 additional subjects are reconstructed using the unrolled model. To validate reconstruction accuracy, global and localized motion estimates from DL model-based 3D iNAVs are compared with those extracted from 3D iNAVs reconstructed with l 1 -ESPIRiT. Then, the high-resolution coronary MRA images motion corrected with autofocusing using the l 1 -ESPIRiT and DL model-based 3D iNAVs are assessed for differences.

RESULTS

3D iNAVs reconstructed using the DL model-based approach and conventional l 1 -ESPIRiT generate similar global and localized motion estimates and provide equivalent coronary image quality. Reconstruction with the unrolled network completes in a fraction of the time compared to CPU and GPU implementations of l 1 -ESPIRiT (20× and 3× speed increases, respectively).

CONCLUSIONS

We have developed a deep neural network architecture to reconstruct undersampled 3D non-Cartesian VD cones iNAVs. Our approach decreases reconstruction time for 3D iNAVs, while preserving the accuracy of nonrigid motion information offered by them for correction.

Machine Learning for Rapid Magnetic Resonance Fingerprinting Tissue Property Quantification

Magnetic Resonance Fingerprinting (MRF) is an MRI-based method that can provide quantitative maps of multiple tissue properties simultaneously from a single rapid acquisition. Tissue property maps are generated by matching the complex signal evolutions collected at the scanner to a dictionary of signals derived using Bloch equation simulations. However, in some circumstances, the process of dictionary generation and signal matching can be time-consuming, reducing the utility of this technique. Recently, several groups have proposed using machine learning to accelerate the extraction of quantitative maps from MRF data. This article will provide an overview of current research that combines MRF and machine learning, as well as present original research demonstrating how machine learning can speed up dictionary generation for cardiac MRF.

Improved pulmonary 129 Xe ventilation imaging via 3D-spiral UTE MRI

PURPOSE

Hyperpolarized ¹²⁹ Xe MRI characterizes regional lung ventilation in a variety of disease populations, with high sensitivity to airway obstruction in early disease. However, ventilation images are usually limited to a single breath-hold and most-often acquired using gradient-recalled echo sequences with thick slices (~10-15 mm), which increases partial-volume effects, limits ability to observe small defects, and suffers from imperfect slice selection. We demonstrate higher-resolution ventilation images, in shorter breath-holds, using FLORET (Fermat Looped ORthogonally Encoded Trajectories), a center-out 3D-spiral UTE sequence.

METHODS

In vivo human adult (N = 4; 2 healthy, 2 with cystic fibrosis) ¹²⁹ Xe images were acquired using 2D gradient-recalled echo, 3D radial, and FLORET. Each sequence was acquired at its highest possible resolution within a 16-second breath-hold with a minimum voxel dimension of 3 mm. Images were compared using ¹²⁹ Xe ventilation defect percentage, SNR, similarity coefficients, and vasculature cross-sections.

RESULTS

The FLORET sequence obtained relative normalized SNR, 40% greater than 2D gradient-recalled echo (P = .012) and 26% greater than 3D radial (P = .067). Moreover, the FLORET images were acquired with 3-fold-higher nominal resolution in a 15% shorter breath-hold. Finally, vasculature was less prominent in FLORET, likely due to diminished susceptibility-induced dephasing at shorter TEs afforded by UTE sequences.

CONCLUSION

The FLORET sequence yields higher SNR for a given resolution with a shorter breath-hold than traditional ventilation imaging techniques. This sequence more accurately measures ventilation abnormalities and enables reduced scan times in patients with poor compliance and severe lung disease.

Prospective GIRF-based RF phase cycling to reduce eddy current-induced steady-state disruption in bSSFP imaging

Purpose

To propose an explicit Balanced steady‐state free precession (bSSFP) signal model that predicts eddy current‐induced steady‐state disruptions and to provide a prospective, practical, and general eddy current compensation method.

Theory and Methods

Gradient impulse response functions (GIRF) were used to simulate trajectory‐specific eddy current‐induced phase errors at the end of a repetition block. These phase errors were included in bloch simulations to establish a bSSFP signal model to predict steady‐state disruptions and their corresponding image artifacts. The signal model was embedded in the MR system and used to compensate the phase errors by prospectively modifying the phase cycling scheme of the RF pulse. The signal model and eddy current compensation method were validated in phantom and in vivo experiments. In addition, the signal model was used to analyze pre‐existing eddy current mitigation methods, such as 2D tiny golden angle radial and 3D paired phase encoded Cartesian acquisitions.

Results

The signal model predicted eddy current‐induced image artifacts, with the zeroth‐order GIRF being the primary factor to predict the steady‐state disruption. Prospective RF phase cycling schemes were automatically computed online and considerably reduced eddy current‐induced image artifacts. The signal model provides a direct relationship for the smoothness of k‐space trajectories, which explains the effectiveness of phase encode pairing and tiny golden angle trajectory.

Conclusions

The proposed signal model can accurately predict eddy current‐induced steady‐state disruptions for bSSFP imaging. The signal model can be used to derive the eddy current‐induced phase errors required for trajectory‐specific RF phase cycling schemes, which considerably reduce eddy current‐induced image artifacts.

Implementation of the FLORET UTE sequence for lung imaging

PURPOSE

Magnetic resonance imaging of lungs is inherently challenging, but it has become more common with the use of UTE sequences and their relative insensitivity to motion. Spiral UTE sequences have been touted recently as having greater k-space sampling efficiencies than radial UTE, but few are designed for the shorter T2 * of the lung. In this study, FLORET (Fermat looped, orthogonally encoded trajectories), a recently developed spiral 3D-UTE sequence designed for the short T2 * species, was implemented in human lungs for the first time and the images were compared with traditional radial UTE images.

METHODS

The FLORET sequence was implemented with parameters optimized for lung imaging on healthy and diseased (cystic fibrosis) subjects. On healthy subjects, radial UTE images (3D-radial and 2D-radial with phase encoding) were acquired for comparison to FLORET. Various metrics including SNR, vasculature contrast, diaphragm sharpness, and parenchymal density ratios were acquired and compared among the separate UTE sequences.

RESULTS

The FLORET sequence performed similarly to traditional radial UTE methods with a much shorter total scan time for fully sampled images (FLORET: 1 minute 55 seconds, 3D-radial: 3 minutes 25 seconds, 2D-radial with phase encoding: 7 minutes 22 seconds). Additionally, the FLORET image obtained on the cystic fibrosis subject resulted in the observation of cystic fibrosis lung pathology similar or superior to that of the other UTE-MRI techniques.

CONCLUSION

The FLORET sequence allows for faster acquisition of high diagnostic-quality lung images and its short T2 * components without sacrificing SNR, image quality, or tissue/disease quantification.

A Fast GPU-optimized 3D MRI Simulator for Arbitrary k-space Sampling

Purpose:

To develop a fast 3D MRI simulator for arbitrary k-space sampling using a graphical processing unit (GPU) and demonstrate its performance by comparing simulation and experimental results in a real MRI system.

Materials and Methods:

A fast 3D MRI simulator using a GeForce GTX 1080 GPU (NVIDIA Corporation, Santa Clara, CA, USA) was developed using C++ and the CUDA 8.0 platform (NVIDIA Corporation). The unique advantage of this simulator was that it could use the same pulse sequence as used in the experiment. The performance of the MRI simulator was measured using two GTX 1080 GPUs and 3D Cones sequences. The MRI simulation results for 3D non-Cartesian sampling trajectories like 3D Cones sequences using a numerical 3D phantom were compared with the experimental results obtained with a real MRI system and a real 3D phantom.

Results:

The performance of the MRI simulator was about 3800–4900 gigaflops for 128- to 4-shot 3D Cones sequences with 256³ voxels, which was about 60% of the performance of the previous MRI simulator optimized for Cartesian sampling calculated for a Cartesian sampling gradient-echo sequence with 256³ voxels. The effects of the static magnetic field inhomogeneity, radio-frequency field inhomogeneity, gradient field nonlinearity, and fast repetition times on the MR images were reproduced in the simulated images as observed in the experimental images.

Conclusion:

The 3D MRI simulator developed for arbitrary k-space sampling optimized using GPUs is a powerful tool for the development and evaluation of advanced imaging sequences including both Cartesian and non-Cartesian k-space sampling.

Bi-exponential 23 Na T2 * component analysis in the human brain

The purpose of this study was to measure the sodium transverse relaxation time T₂ * in the healthy human brain. Five healthy subjects were scanned with 18 echo times (TEs) as short as 0.17 ms. T₂ * values were fitted on a voxel-by-voxel basis using a bi-exponential model. Data were also analysed using a continuous distribution fit with a region of interest-based inverse Laplace transform. Average T₂ * values were 3.4 ± 0.2 ms and 23.5 ± 1.8 ms in white matter (WM) for the short and long components, respectively, and 3.9 ± 0.5 ms and 26.3 ± 2.6 ms in grey matter (GM) for the short and long components, respectively, using the bi-exponential model. Continuous distribution fits yielded results of 3.1 ± 0.3 ms and 18.8 ± 3.2 ms in WM for the short and long components, respectively, and 2.9 ± 0.4 ms and 17.2 ± 2 ms in GM for the short and long components, respectively. ²³ Na T₂ * values of the brain for the short and long components for various anatomical locations using ultra-short TEs are presented for the first time.

Volumetric MRI thermometry using a three-dimensional stack-of-stars echo-planar imaging pulse sequence

PURPOSE

To measure temperature over a large brain volume with fine spatiotemporal resolution.

METHODS

A three-dimensional stack-of-stars echo-planar imaging sequence combining echo-planar imaging and radial sampling with golden angle spacing was implemented at 3T for proton resonance frequency-shift temperature imaging. The sequence acquires a 188x188x43 image matrix with 1.5x1.5x2.75 mm3 spatial resolution. Temperature maps were reconstructed using sensitivity encoding (SENSE) image reconstruction followed by the image domain hybrid method, and using the k-space hybrid method. In vivo temperature maps were acquired without heating to measure temperature precision in the brain, and in a phantom during high-intensity focused ultrasound sonication.

RESULTS

In vivo temperature standard deviation was less than 1°C at dynamic scan times down to 0.75 s. For a given frame rate, scanning at a minimum repetition time (TR) with minimum acceleration yielded the lowest standard deviation. With frame rates around 3 s, the scan was tolerant to a small number of receive coils, and temperature standard deviation was 48% higher than a standard two-dimensional Fourier transform temperature mapping scan, but provided whole-brain coverage. Phantom temperature maps with no visible aliasing were produced for dynamic scan times as short as 0.38 s. k-Space hybrid reconstructions were more tolerant to acceleration.

CONCLUSION

Three-dimensional stack-of-stars echo-planar imaging temperature mapping provides volumetric brain coverage and fine spatiotemporal resolution. Magn Reson Med 79:2003-2013, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Pulmonary artery imaging under free-breathing using golden-angle radial bSSFP MRI: a proof of concept

PURPOSE

To evaluate the feasibility of an improved motion and flow robust methodology for imaging the pulmonary vasculature using non-contrast-enhanced, free-breathing, golden-angle radial MRI.

METHODS

Healthy volunteers (n = 10, age 46 ± 11 years, 50% female) and patients (n = 2, ages 27 and 84, both female) were imaged at 1.5 T using a Cartesian and golden-angle radial 2D balanced SSFP pulse sequence. The acquisitions were made under free breathing without contrast agent enhancement. The radial acquisitions were reconstructed at 3 temporal footprints. All series were scored from 1 to 5 for perceived diagnostic quality, artifact level, and vessel sharpness in multiple anatomical locations. In addition, vessel sharpness and blood-to-blood clot contrast were measured.

RESULTS

Quantitative measurements showed higher vessel sharpness for golden-angle radial (n = 76, 0.79 ± 0.11 versus 0.71 ± 0.16, p < .05). Blood-to-blood clot contrast was found to be 23% higher in golden-angle radial in the 2 patients. At comparable temporal footprints, golden-angle radial was scored higher for diagnostic quality (mean ± SD, 2.3 ± 0.7 versus 2.2 ± 0.6, p < .01) and vessel sharpness (2.2 ± 0.8 versus 2.1 ± 0.5, p < .01), whereas the artifact level did not differ (3.0 ± 0.9 versus 3.0 ± 1.0, p = .80). The ability to retrospectively choose a temporal resolution and perform sliding-window reconstructions was demonstrated in patients.

CONCLUSION

In pulmonary artery imaging, the motion and flow robustness of a radial trajectory does both improve image quality over Cartesian trajectory in healthy volunteers, and allows for flexible selection of temporal footprints and the ability to perform real-time sliding window reconstructions, which could potentially provide further diagnostic insight.

Optimization and validation of accelerated golden-angle radial sparse MRI reconstruction with self-calibrating GRAPPA operator gridding

PURPOSE

Golden-angle radial sparse parallel (GRASP) MRI reconstruction requires gridding and regridding to transform data between radial and Cartesian k-space. These operations are repeatedly performed in each iteration, which makes the reconstruction computationally demanding. This work aimed to accelerate GRASP reconstruction using self-calibrating GRAPPA operator gridding (GROG) and to validate its performance in clinical imaging.

METHODS

GROG is an alternative gridding approach based on parallel imaging, in which k-space data acquired on a non-Cartesian grid are shifted onto a Cartesian k-space grid using information from multicoil arrays. For iterative non-Cartesian image reconstruction, GROG is performed only once as a preprocessing step. Therefore, the subsequent iterative reconstruction can be performed directly in Cartesian space, which significantly reduces computational burden. Here, a framework combining GROG with GRASP (GROG-GRASP) is first optimized and then compared with standard GRASP reconstruction in 22 prostate patients.

RESULTS

GROG-GRASP achieved approximately 4.2-fold reduction in reconstruction time compared with GRASP (∼333 min versus ∼78 min) while maintaining image quality (structural similarity index ≈ 0.97 and root mean square error ≈ 0.007). Visual image quality assessment by two experienced radiologists did not show significant differences between the two reconstruction schemes. With a graphics processing unit implementation, image reconstruction time can be further reduced to approximately 14 min.

CONCLUSION

The GRASP reconstruction can be substantially accelerated using GROG. This framework is promising toward broader clinical application of GRASP and other iterative non-Cartesian reconstruction methods. Magn Reson Med 80:286-293, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Rapid compressed sensing reconstruction of 3D non-Cartesian MRI

PURPOSE

Conventional non-Cartesian compressed sensing requires multiple nonuniform Fourier transforms every iteration, which is computationally expensive. Accordingly, time-consuming reconstructions have slowed the adoption of undersampled 3D non-Cartesian acquisitions into clinical protocols. In this work we investigate several approaches to minimize reconstruction times without sacrificing accuracy.

METHODS

The reconstruction problem can be reformatted to exploit the Toeplitz structure of matrices that are evaluated every iteration, but it requires larger oversampling than what is strictly required by nonuniform Fourier transforms. Accordingly, we investigate relative speeds of the two approaches for various nonuniform Fourier transform kernel sizes and oversampling for both GPU and CPU implementations. Second, we introduce a method to minimize matrix sizes by estimating the image support. Finally, density compensation weights have been used as a preconditioning matrix to improve convergence, but this increases noise. We propose a more general approach to preconditioning that allows a trade-off between accuracy and convergence speed.

RESULTS

When using a GPU, the Toeplitz approach was faster for all practical parameters. Second, it was found that properly accounting for image support can prevent aliasing errors with minimal impact on reconstruction time. Third, the proposed preconditioning scheme improved convergence rates by an order of magnitude with negligible impact on noise.

CONCLUSION

With the proposed methods, 3D non-Cartesian compressed sensing with clinically relevant reconstruction times (<2 min) is feasible using practical computer resources. Magn Reson Med 79:2685-2692, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Technical Note: Evaluation of pre-reconstruction interpolation methods for iterative reconstruction of radial k-space data

PURPOSE

To evaluate the use of three different pre-reconstruction interpolation methods to convert non-Cartesian k-space data to Cartesian samples such that iterative reconstructions can be performed more simply and more rapidly.

METHODS

Phantom as well as cardiac perfusion radial datasets were reconstructed by four different methods. Three of the methods used pre-reconstruction interpolation once followed by a fast Fourier transform (FFT) at each iteration. The methods were: bilinear interpolation of nearest-neighbor points (BINN), 3-point interpolation, and a multi-coil interpolator called GRAPPA Operator Gridding (GROG). The fourth method performed a full non-Uniform FFT (NUFFT) at each iteration. An iterative reconstruction with spatiotemporal total variation constraints was used with each method. Differences in the images were quantified and compared.

RESULTS

The GROG multicoil interpolation, the 3-point interpolation, and the NUFFT-at-each-iteration approaches produced high quality images compared to BINN, with the GROG-derived images having the fewest streaks among the three preinterpolation approaches. However, all reconstruction methods produced approximately equal results when applied to perfusion quantitation tasks. Pre-reconstruction interpolation gave approximately an 83% reduction in reconstruction time.

CONCLUSION

Image quality suffers little from using a pre-reconstruction interpolation approach compared to the more accurate NUFFT-based approach. GROG-based pre-reconstruction interpolation appears to offer the best compromise by using multicoil information to perform the interpolation to Cartesian sample points prior to image reconstruction. Speed gains depend on the implementation and relatively standard optimizations on a MATLAB platform result in preinterpolation speedups of ~ 6 compared to using NUFFT at every iteration, reducing the reconstruction time from around 42 min to 7 min.

Accelerated mapping of magnetic susceptibility using 3D planes-on-a-paddlewheel (POP) EPI at ultra-high field strength

With the advent of ultra-high field MRI scanners in clinical research, susceptibility based MRI has recently gained increasing interest because of its potential to assess subtle tissue changes underlying neurological pathologies/disorders. Conventional, but rather slow, three-dimensional (3D) spoiled gradient-echo (GRE) sequences are typically employed to assess the susceptibility of tissue. 3D echo-planar imaging (EPI) represents a fast alternative but generally comes with echo-time restrictions, geometrical distortions and signal dropouts that can become severe at ultra-high fields. In this work we assess quantitative susceptibility mapping (QSM) at 7 T using non-Cartesian 3D EPI with a planes-on-a-paddlewheel (POP) trajectory, which is created by rotating a standard EPI readout train around its own phase encoding axis. We show that the threefold accelerated non-Cartesian 3D POP EPI sequence enables very fast, whole brain susceptibility mapping at an isotropic resolution of 1 mm and that the high image quality has sufficient signal-to-noise ratio in the phase data for reliable QSM processing. The susceptibility maps obtained were comparable with regard to QSM values and geometric distortions to those calculated from a conventional 4 min 3D GRE scan using the same QSM processing pipeline. Copyright © 2016 John Wiley & Sons, Ltd.

3D-accelerated, stack-of-spirals acquisitions and reconstruction of arterial spin labeling MRI

PURPOSE

The goal of this study was to develop a 3D acceleration and reconstruction method to improve image quality and resolution of background-suppressed arterial spin-labeled perfusion MRI.

METHODS

Accelerated acquisition was implemented in all three k-space dimensions in a stack-of-spirals readout using variable density spirals and partition undersampling. A single 3D self-consistent parallel imaging (SPIRiT) kernel was calibrated and iteratively applied to reconstruct each imaging volume. Whole-brain (including cerebellum) perfusion imaging was obtained at 3-mm isotropic resolution (nominal) using single- and 2-shot acquisitions and at 2-mm isotropic resolution (nominal) using four-shot acquisitions, achieving effective acceleration factors between 5.5 and 6.6. The signal-to-noise (SNR) performance of 3D SPIRiT was evaluated. The temporal SNR (tSNR) of the cerebral blood flow (CBF) maps and the gray/white matter CBF ratios were quantified.

RESULTS

The readout of the arterial spin labeling (ASL) sequence was significantly shortened with acceleration. The CBF values were consistent between accelerated and fully sampled ASL. With shorter spiral interleaves and shorter echo trains, the accelerated images demonstrated reduced blurring and signal dropout in regions with high susceptibility gradients, resulting in improved image quality and increased gray/white matter CBF ratios. The shortened readout was accompanied by a corresponding decrease in tSNR.

CONCLUSION

The 3D acceleration and reconstruction allow a rapid whole-brain readout that improved the quality of ASL perfusion imaging. Magn Reson Med 78:1405-1419, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Three-dimensional ultrashort echo-time imaging using a FLORET trajectory

PURPOSE

Three-dimensional ultrashort echo-time (UTE) imaging commonly makes use of an isotropic 3D radial projection acquisition. The FLORET sequence is proposed and evaluated as a more efficient alternative.

METHODS

The properties of the FLORET trajectory are contrasted with those of a 3D radial projection trajectory. The theoretical advantages of FLORET, including greater sampling and SNR efficiency, are evaluated based upon experimental data. The effect of T2* decay on FLORET is analyzed in comparison to the 3D radial, Cones, and Density Adapted Radial trajectories. FLORET UTE image quality is compared with 3D radial UTE image quality.

RESULTS

FLORET is shown to have several advantages over 3D radial acquisitions with respect to image quality, scan time, signal-to-noise, and off-resonance blurring for UTE data. The signal and resolution losses from T2* decay for a FLORET acquisition are shown to be comparable to those of Density Adapted Radial and Density Compensated Cones trajectories.

CONCLUSION

The FLORET sequence is recommended as an alternative to 3D radial projection sequences for musculoskeletal UTE imaging as well as other UTE applications that accommodate modest to long per shot sampling times. FLORET is not recommended for imaging extremely short T₂ species such as dentin. Magn Reson Med 78:1038-1049, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Self-gated MRI of multiple beat morphologies in the presence of arrhythmias

PURPOSE

Develop self-gated MRI for distinct heartbeat morphologies in subjects with arrhythmias.

METHODS

Golden angle radial data was obtained in seven sinus and eight arrhythmias subjects. An image-based cardiac navigator was derived from single-shot images, distinct beat types were identified, and images were reconstructed for repeated morphologies. Image sharpness, contrast, and volume variation were quantified and compared with self-gated MRI. Images were scored for image quality and artifacts. Hemodynamic parameters were computed for each distinct beat morphology in bigeminy and trigeminy subjects and for sinus beats in patients with infrequent premature ventricular contractions.

RESULTS

Images of distinct beat types were reconstructed except for two patients with infrequent premature ventricular contractions. Image contrast and sharpness were similar to sinus self-gated images (contrast = 0.45 ± 0.13 and 0.43 ± 0.15; sharpness = 0.21 ± 0.11 and 0.20 ± 0.05). Visual scoring was highest in self-gated images (4.1 ± 0.3) compared with real-time (3.9 ± 0.4) and ECG-gated cine (3.4 ± 1.5). ECG-gated cine had less artifacts than self-gating (2.3 ± 0.7 and 2.1 ± 0.2), but was affected by misgating in two subjects. Among arrhythmia subjects, post-extrasystole/sinus (58.1 ± 8.6 mL) and interrupted sinus (61.4 ± 5.9 mL) stroke volume was higher than extrasystole (32.0 ± 16.5 mL; P < 0.02).

CONCLUSION

Self-gated imaging can reconstruct images during ectopy and allowed for quantification of hemodynamic function of different beat morphologies. Magn Reson Med 78:678-688, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

B0 mapping using rewinding trajectories (BMART)

PURPOSE

To create a B₀ map and correct for off-resonance with minimal scan time increase for two-dimensional (2D) or 3D non-Cartesian acquisitions.

METHODS

Rewinding trajectories that bring the zeroth gradient moment to zero every repetition time (TR) were used to estimate the off-resonance with a center-out 3D cones trajectory, which required an increase in the minimum TR by 5%. The off-resonance estimation and correction was implemented using an algorithm based on binning and object-domain phase correction. B₀ maps using BMART (B₀ mapping using rewinding trajectories) were compared to maps obtained using separate scans with multiple echo time (TE) in a phantom and human brain.

RESULTS

Excellent agreement between BMART and the multiple-TE method were observed, and images corrected with BMART were deblurred.

CONCLUSION

BMART can correct for off-resonance without requiring an additional scan, and can be easily applied to center-out or projection trajectories (2D or 3D). Magn Reson Med 78:664-669, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Undersampled linogram trajectory for fast imaging (ULTI): experiments at 3 T and 7 T

In this study, the performance of linogram acquisition was investigated for the reconstruction of images from undersampled data using parallel imaging methods. The point spread function (PSF) of linogram sampling was analyzed for image sharpness and artifacts. Generalized auto-calibrating partially parallel acquisition was implemented for this new sampling scheme, and images were reconstructed with high acceleration rates. The results were compared with conventional radial sampling methods using simulations and phantom experiments at 3 T. Additionally, a human volunteer was scanned at 7 T. The results demonstrated that the PSF was sharper and the mean artifact power was lower for linogram sampling compared with radial sampling. Results of simulations and phantom experiments were in accord with the findings of the PSF analysis. In simulations, errors in the reconstructed images were lower for linogram sampling. In phantom experiments, fine details and sharp edges were preserved for linogram sampling, while details were blurred for radial sampling. The in vivo human study demonstrated that linogram sampling could provide high quality images of anatomy, even at high acceleration rates. Linogram sampling not only possesses the advantages of radial sampling, such as reduced sensitivity to motion and higher acceleration rates, but it also provides sharper images with fewer artifacts. Moreover, it is less prone to off-resonance artifacts compared with radial sampling. Copyright © 2016 John Wiley & Sons, Ltd.

Ungated radial quiescent-inflow single-shot (UnQISS) magnetic resonance angiography using optimized azimuthal equidistant projections

PURPOSE

We hypothesized that non-contrast-enhanced MR angiography (NEMRA) could be performed without cardiac gating by using a variant of the quiescent-inflow single-shot (QISS) technique.

METHODS

Ungated QISS (UnQISS) MRA was evaluated in eight patients with peripheral arterial disease at 1.5T. The radial acquisition used optimized azimuthal equidistant projections, a long quiescent inflow time (1200 ms) to ensure replenishment of saturated in-plane spins irrespective of the cardiac phase, and a lengthy readout (1200 ms) so that a complete cardiac cycle was sampled for each slice. Venous and background tissue suppression was obtained using frequency-offset-corrected inversion radiofrequency pulses.

RESULTS

Scan time for UnQISS was 15.4 min for an eight-station whole-leg acquisition. The appearance of UnQISS MRA acquired using the body coil was comparable to electrocardiographic-gated QISS MRA using phased array coils. A small radial view angle increment minimized eddy current-related artifacts, whereas image quality was inferior with a golden view angle radial increment or Cartesian trajectory. In patient studies, ≥50% stenoses were consistently detected.

CONCLUSION

Using UnQISS, peripheral NEMRA can be performed without the need for cardiac gating. The use of fixed imaging parameters and body coil for signal reception further simplifies the scan procedure.

Non-Cartesian parallel imaging reconstruction

Non-Cartesian parallel imaging has played an important role in reducing data acquisition time in MRI. The use of non-Cartesian trajectories can enable more efficient coverage of k-space, which can be leveraged to reduce scan times. These trajectories can be undersampled to achieve even faster scan times, but the resulting images may contain aliasing artifacts. Just as Cartesian parallel imaging can be used to reconstruct images from undersampled Cartesian data, non-Cartesian parallel imaging methods can mitigate aliasing artifacts by using additional spatial encoding information in the form of the nonhomogeneous sensitivities of multi-coil phased arrays. This review will begin with an overview of non-Cartesian k-space trajectories and their sampling properties, followed by an in-depth discussion of several selected non-Cartesian parallel imaging algorithms. Three representative non-Cartesian parallel imaging methods will be described, including Conjugate Gradient SENSE (CG SENSE), non-Cartesian generalized autocalibrating partially parallel acquisition (GRAPPA), and Iterative Self-Consistent Parallel Imaging Reconstruction (SPIRiT). After a discussion of these three techniques, several potential promising clinical applications of non-Cartesian parallel imaging will be covered.

More IMPATIENT: A Gridding-Accelerated Toeplitz-based Strategy for Non-Cartesian High-Resolution 3D MRI on GPUs

Several recent methods have been proposed to obtain significant speed-ups in MRI image reconstruction by leveraging the computational power of GPUs. Previously, we implemented a GPU-based image reconstruction technique called the Illinois Massively Parallel Acquisition Toolkit for Image reconstruction with ENhanced Throughput in MRI (IMPATIENT MRI) for reconstructing data collected along arbitrary 3D trajectories. In this paper, we improve IMPATIENT by removing computational bottlenecks by using a gridding approach to accelerate the computation of various data structures needed by the previous routine. Further, we enhance the routine with capabilities for off-resonance correction and multi-sensor parallel imaging reconstruction. Through implementation of optimized gridding into our iterative reconstruction scheme, speed-ups of more than a factor of 200 are provided in the improved GPU implementation compared to the previous accelerated GPU code.

Volumetric late gadolinium-enhanced myocardial imaging with retrospective inversion time selection

PURPOSE

To develop and validate a novel free-breathing 3D radial late gadolinium-enhanced magnetic resonance imaging technique (3D LGE-MRI) with isotropic resolution and retrospective inversion time (TI) selection for myocardial viability imaging.

MATERIALS AND METHODS

The 3D radial LGE-MRI method featuring an interleaved and bit-reversed radial k-space trajectory was evaluated in 12 subjects that also had clinical breath-hold Cartesian 2D LGE-MRI. The 3D LGE-MRI acquisition requires a predicted TI and a user-controlled data acquisition window that determines the sampling width around the predicted TI. Sliding window reconstructions with update rates of 1× the repetition time (TR) allow for a user selectable TI to obtain the maximum nulling of the myocardium. The retrospective nature of the acquisition allows the user to choose from a range of possible TI times centered on the expected TI. Those projections most corrupted by respiratory motion, as determined by a respiratory bellows signal, were resampled according to the diminishing variance algorithm. The quality of the left ventricular myocardial nulling on the 3D LGE-MRI and 2D LGE-MRI was assessed using a 4-point Likert scale by two experienced radiologists. Comparison of image quality scores for the two methods was performed using generalized estimating equations.

RESULTS

All 3D LGE-MRI cases produced similar nulling of myocardial signal as the 2D LGE-MRI. The image quality of myocardial nulling was not significantly different between the two acquisitions (mean nulling of 3.4 for 2D vs. 3.1 for 3D, and P = 0.0645). The average absolute deviation from mean scores was also not determined to be statistically significant (1.8 for 2D and 0.4 for 3D and P = 0.1673). Total acquisition time was ∼9 minutes for 3D LGE-MRI with voxel sizes ranging from 1.6(3) to 2.0(3) mm(3) . Conversely, the total imaging time was twice as long for the 2D DCE-MRI (>17 minutes) with an eight times larger voxel size of 1.4 × 2.2 × 7.0 mm.

CONCLUSION

The 3D LGE-MRI technique demonstrated in this study is a promising alternative for the assessment of myocardial viability in patients who have difficulty sustaining breath-holds for the clinical standard 2D LGE-MRI.