Rational approximation of golden angles: Accelerated reconstructions for radial MRI

PURPOSE

To develop a generic radial sampling scheme that combines the advantages of golden ratio sampling with simplicity of equidistant angular patterns. The irrational angle between consecutive spokes in golden ratio-based sampling schemes enables a flexible retrospective choice of temporal resolution, while preserving good coverage of k-space for each individual bin. Nevertheless, irrational increments prohibit precomputation of the point-spread function (PSF), can lead to numerical problems, and require more complex processing steps. To avoid these problems, a new sampling scheme based on a rational approximation of golden angles (RAGA) is developed.

METHODS

The theoretical properties of RAGA sampling are mathematically derived. Sidelobe-to-peak ratios (SPR) are numerically computed and compared to the corresponding golden ratio sampling schemes. The sampling scheme is implemented in the BART toolbox and in a radial gradient-echo sequence. Feasibility is shown for quantitative imaging in a phantom and a cardiac scan of a healthy volunteer.

RESULTS

RAGA sampling can accurately approximate golden ratio sampling and has almost identical PSF and SPR. In contrast to golden ratio sampling, each frame can be reconstructed with the same equidistant trajectory using different sampling masks, and the angle of each acquired spoke can be encoded as a small index, which simplifies processing of the acquired data.

CONCLUSION

RAGA sampling provides the advantages of golden ratio sampling while simplifying data processing, rendering it a valuable tool for dynamic and quantitative MRI.

Deep learning for accelerated and robust MRI reconstruction

Deep learning (DL) has recently emerged as a pivotal technology for enhancing magnetic resonance imaging (MRI), a critical tool in diagnostic radiology. This review paper provides a comprehensive overview of recent advances in DL for MRI reconstruction, and focuses on various DL approaches and architectures designed to improve image quality, accelerate scans, and address data-related challenges. It explores end-to-end neural networks, pre-trained and generative models, and self-supervised methods, and highlights their contributions to overcoming traditional MRI limitations. It also discusses the role of DL in optimizing acquisition protocols, enhancing robustness against distribution shifts, and tackling biases. Drawing on the extensive literature and practical insights, it outlines current successes, limitations, and future directions for leveraging DL in MRI reconstruction, while emphasizing the potential of DL to significantly impact clinical imaging practices.

Quantitative MR Image Reconstruction Using Parameter-Specific Dictionary Learning With Adaptive Dictionary-Size and Sparsity-Level Choice

OBJECTIVE

We propose a method for the reconstruction of parameter-maps in Quantitative Magnetic Resonance Imaging (QMRI).

METHODS

Because different quantitative parameter-maps differ from each other in terms of local features, we propose a method where the employed dictionary learning (DL) and sparse coding (SC) algorithms automatically estimate the optimal dictionary-size and sparsity level separately for each parameter-map. We evaluated the method on a T1-mapping QMRI problem in the brain using the BrainWeb data as well as in-vivo brain images acquired on an ultra-high field 7 T scanner. We compared it to a model-based acceleration for parameter mapping (MAP) approach, other sparsity-based methods using total variation (TV), Wavelets (Wl), and Shearlets (Sh) to a method which uses DL and SC to reconstruct qualitative images, followed by a non-linear (DL+Fit).

RESULTS

Our algorithm surpasses MAP, TV, Wl, and Sh in terms of RMSE and PSNR. It yields better or comparable results to DL+Fit by additionally significantly accelerating the reconstruction by a factor of approximately seven.

CONCLUSION

The proposed method outperforms the reported methods of comparison and yields accurate T1-maps. Although presented for T1-mapping in the brain, our method's structure is general and thus most probably also applicable for the the reconstruction of other quantitative parameters in other organs.

SIGNIFICANCE

From a clinical perspective, the obtained T1-maps could be utilized to differentiate between healthy subjects and patients with Alzheimer's disease. From a technical perspective, the proposed unsupervised method could be employed to obtain ground-truth data for the development of data-driven methods based on supervised learning.

A low-rank deep image prior reconstruction for free-breathing ungated spiral functional CMR at 0.55 T and 1.5 T

OBJECTIVE

This study combines a deep image prior with low-rank subspace modeling to enable real-time (free-breathing and ungated) functional cardiac imaging on a commercial 0.55 T scanner.

MATERIALS AND METHODS

The proposed low-rank deep image prior (LR-DIP) uses two u-nets to generate spatial and temporal basis functions that are combined to yield dynamic images, with no need for additional training data. Simulations and scans in 13 healthy subjects were performed at 0.55 T and 1.5 T using a golden angle spiral bSSFP sequence with images reconstructed using [Formula: see text]-ESPIRiT, low-rank plus sparse (L + S) matrix completion, and LR-DIP. Cartesian breath-held ECG-gated cine images were acquired for reference at 1.5 T. Two cardiothoracic radiologists rated images on a 1-5 scale for various categories, and LV function measurements were compared.

RESULTS

LR-DIP yielded the lowest errors in simulations, especially at high acceleration factors (R [Formula: see text] 8). LR-DIP ejection fraction measurements agreed with 1.5 T reference values (mean bias - 0.3% at 0.55 T and - 0.2% at 1.5 T). Compared to reference images, LR-DIP images received similar ratings at 1.5 T (all categories above 3.9) and slightly lower at 0.55 T (above 3.4).

CONCLUSION

Feasibility of real-time functional cardiac imaging using a low-rank deep image prior reconstruction was demonstrated in healthy subjects on a commercial 0.55 T scanner.

Free-running 3D whole-heart T1 and T2 mapping and cine MRI using low-rank reconstruction with non-rigid cardiac motion correction

PURPOSE

To introduce non-rigid cardiac motion correction into a novel free-running framework for the simultaneous acquisition of 3D whole-heart myocardial <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>1</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_1 $$</mml:annotation></mml:semantics> </mml:math> and <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>2</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_2 $$</mml:annotation></mml:semantics> </mml:math> maps and cine images, enabling a <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mo>∼</mml:mo></mml:mrow> <mml:annotation>$$ \sim $$</mml:annotation></mml:semantics> </mml:math> 3-min scan.

METHODS

Data were acquired using a free-running 3D golden-angle radial readout interleaved with inversion recovery and <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>2</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_2 $$</mml:annotation></mml:semantics> </mml:math> -preparation pulses. After correction for translational respiratory motion, non-rigid cardiac-motion-corrected reconstruction with dictionary-based low-rank compression and patch-based regularization enabled 3D whole-heart <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>1</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_1 $$</mml:annotation></mml:semantics> </mml:math> and <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>2</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_2 $$</mml:annotation></mml:semantics> </mml:math> mapping at any given cardiac phase as well as whole-heart cardiac cine imaging. The framework was validated and compared with established methods in 11 healthy subjects.

RESULTS

Good quality 3D <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>1</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_1 $$</mml:annotation></mml:semantics> </mml:math> and <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>2</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_2 $$</mml:annotation></mml:semantics> </mml:math> maps and cine images were reconstructed for all subjects. Septal <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>1</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_1 $$</mml:annotation></mml:semantics> </mml:math> values using the proposed approach ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mn>1200</mml:mn> <mml:mo>±</mml:mo> <mml:mn>50</mml:mn></mml:mrow> <mml:annotation>$$ 1200\pm 50 $$</mml:annotation></mml:semantics> </mml:math> ms) were higher than those from a 2D MOLLI sequence ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mn>1063</mml:mn> <mml:mo>±</mml:mo> <mml:mn>33</mml:mn></mml:mrow> <mml:annotation>$$ 1063\pm 33 $$</mml:annotation></mml:semantics> </mml:math> ms), which is known to underestimate <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>1</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_1 $$</mml:annotation></mml:semantics> </mml:math> , while <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>2</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_2 $$</mml:annotation></mml:semantics> </mml:math> values from the proposed approach ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mn>51</mml:mn> <mml:mo>±</mml:mo> <mml:mn>4</mml:mn></mml:mrow> <mml:annotation>$$ 51\pm 4 $$</mml:annotation></mml:semantics> </mml:math> ms) were in good agreement with those from a 2D GraSE sequence ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mn>51</mml:mn> <mml:mo>±</mml:mo> <mml:mn>2</mml:mn></mml:mrow> <mml:annotation>$$ 51\pm 2 $$</mml:annotation></mml:semantics> </mml:math> ms).

CONCLUSION

The proposed technique provides 3D <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>1</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_1 $$</mml:annotation></mml:semantics> </mml:math> and <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics> <mml:mrow> <mml:msub><mml:mrow><mml:mi>T</mml:mi></mml:mrow> <mml:mrow><mml:mn>2</mml:mn></mml:mrow> </mml:msub> </mml:mrow> <mml:annotation>$$ {T}_2 $$</mml:annotation></mml:semantics> </mml:math> maps and cine images with isotropic spatial resolution in a single <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:semantics><mml:mrow><mml:mo>∼</mml:mo></mml:mrow> <mml:annotation>$$ \sim $$</mml:annotation></mml:semantics> </mml:math> 3.3-min scan.

Evaluation of combined late gadolinium-enhancement and functional cardiac magnetic resonance imaging using spiral real-time acquisition

The purpose of the current study was to implement and validate joint real-time acquisition of functional and late gadolinium-enhancement (LGE) cardiac magnetic resonance (MR) images during free breathing. Inversion recovery cardiac real-time images with a temporal resolution of 50 ms were acquired using a spiral trajectory (IR-CRISPI) with a pre-emphasis based on the gradient system transfer function during free breathing. Functional and LGE cardiac MR images were reconstructed using a low-rank plus sparse model. Late gadolinium-enhancement appearance, image quality, and functional parameters of IR-CRISPI were compared with clinical standard balanced steady-state free precession breath-hold techniques in 10 patients. The acquisition of IR-CRISPI in free breathing of the entire left ventricle took 97 s on average. Bland-Altman analysis and Wilcoxon tests showed a higher artifact level for the breath-hold technique (p = 0.003), especially for arrhythmic patients or patients with dyspnea, but an increased noise level for IR-CRISPI of the LGE images (p = 0.01). The estimated transmural extent of the enhancement differed by not more than 25% and did not show a significant bias between the techniques (p = 0.50). The ascertained functional parameters were similar for the breath-hold technique and IR-CRISPI, that is, with a minor, nonsignificant (p = 0.16) mean difference of the ejection fraction of 2.3% and a 95% confidence interval from -4.8% to 9.4%. IR-CRISPI enables joint functional and LGE imaging in free breathing with good image quality but distinctly shorter scan times in comparison with breath-hold techniques.

A Self-Supervised Deep Learning Reconstruction for Shortening the Breathhold and Acquisition Window in Cardiac Magnetic Resonance Fingerprinting

The aim of this study is to shorten the breathhold and diastolic acquisition window in cardiac magnetic resonance fingerprinting (MRF) for simultaneous T1, T2, and proton spin density (M0) mapping to improve scan efficiency and reduce motion artifacts. To this end, a novel reconstruction was developed that combines low-rank subspace modeling with a deep image prior, termed DIP-MRF. A system of neural networks is used to generate spatial basis images and quantitative tissue property maps, with training performed using only the undersampled k-space measurements from the current scan. This approach avoids difficulties with obtaining in vivo MRF training data, as training is performed de novo for each acquisition. Calculation of the forward model during training is accelerated by using GRAPPA operator gridding to shift spiral k-space data to Cartesian grid points, and by using a neural network to rapidly generate fingerprints in place of a Bloch equation simulation. DIP-MRF was evaluated in simulations and at 1.5 T in a standardized phantom, 18 healthy subjects, and 10 patients with suspected cardiomyopathy. In addition to conventional mapping, two cardiac MRF sequences were acquired, one with a 15-heartbeat(HB) breathhold and 254 ms acquisition window, and one with a 5HB breathhold and 150 ms acquisition window. In simulations, DIP-MRF yielded decreased nRMSE compared to dictionary matching and a sparse and locally low rank (SLLR-MRF) reconstruction. Strong correlation (R2 &gt; 0.999) with T1 and T2 reference values was observed in the phantom using the 5HB/150 ms scan with DIP-MRF. DIP-MRF provided better suppression of noise and aliasing artifacts in vivo, especially for the 5HB/150 ms scan, and lower intersubject and intrasubject variability compared to dictionary matching and SLLR-MRF. Furthermore, it yielded a better agreement between myocardial T1 and T2 from 15HB/254 ms and 5HB/150 ms MRF scans, with a bias of −9 ms for T1 and 2 ms for T2. In summary, this study introduces an extension of the deep image prior framework for cardiac MRF tissue property mapping, which does not require pre-training with in vivo scans, and has the potential to reduce motion artifacts by enabling a shortened breathhold and acquisition window.

Iterative static field map estimation for off-resonance correction in non-Cartesian susceptibility weighted imaging

Purpose

Patient‐induced inhomogeneities in the magnetic field cause distortions and blurring during acquisitions with long readouts such as in susceptibility‐weighted imaging (SWI). Most correction methods require collecting an additional ΔB0 field map to remove these artifacts.

Theory

The static ΔB0 field map can be approximated with an acceptable error directly from a single echo acquisition in SWI. The main component of the observed phase is linearly related to ΔB0 and the echo time (TE), and the relative impact of non‐ ΔB0 terms becomes insignificant with TE >20 ms at 3 T for a well‐tuned system.

Methods

The main step is to combine and unfold the multi‐channel phase maps wrapped many times, and several competing algorithms are compared for this purpose. Four in vivo brain data sets collected using the recently proposed 3D spreading projection algorithm for rapid k‐space sampling (SPARKLING) readouts are used to assess the proposed method.

Results

The estimated 3D field maps generated with a 0.6 mm isotropic spatial resolution provide overall similar off‐resonance corrections compared to reference corrections based on an external ΔB0 acquisitions, and even improved for 2 of 4 individuals. Although a small estimation error is expected, no aftermath was observed in the proposed corrections, whereas degradations were observed in the references.

Conclusion

A static ΔB0 field map estimation method was proposed to take advantage of acquisitions with long echo times, and outperformed the reference technique based on an external field map. The difference can be attributed to an inherent robustness to mismatches between volumes and external ΔB0 maps, and diverse other sources investigated.

Click here for author‐reader discussions

Model-based dynamic off-resonance correction for improved accelerated fMRI in awake behaving nonhuman primates

Purpose

To estimate dynamic off‐resonance due to vigorous body motion in accelerated fMRI of awake behaving nonhuman primates (NHPs) using the echo‐planar imaging reference navigator, in order to attenuate the effects of time‐varying off‐resonance on the reconstruction.

Methods

In NHP fMRI, the animal’s head is usually head‐posted, and the dynamic off‐resonance is mainly caused by motion in body parts that are distant from the brain and have low spatial frequency. Hence, off‐resonance at each frame can be approximated as a spatially linear perturbation of the off‐resonance at a reference frame, and is manifested as a relative linear shift in k‐space. Using GRAPPA operators, we estimated these shifts by comparing the navigator at each time frame with that at the reference frame. Estimated shifts were then used to correct the data at each frame. The proposed method was evaluated in phantom scans, simulations, and in vivo data.

Results

The proposed method is shown to successfully estimate spatially low‐order dynamic off‐resonance perturbations, including induced linear off‐resonance perturbations in phantoms, and is able to correct retrospectively corrupted data in simulations. Finally, it is shown to reduce ghosting artifacts and geometric distortions by up to 20% in simultaneous multislice in vivo acquisitions in awake‐behaving NHPs.

Conclusion

A method is proposed that does not need sequence modification or extra acquisitions and makes accelerated awake behaving NHP imaging more robust and reliable, reducing the gap between what is possible with NHP protocols and state‐of‐the‐art human imaging.

Real-Time Magnetic Resonance Imaging

Real-time magnetic resonance imaging (RT-MRI) allows for imaging dynamic processes as they occur, without relying on any repetition or synchronization. This is made possible by modern MRI technology such as fast-switching gradients and parallel imaging. It is compatible with many (but not all) MRI sequences, including spoiled gradient echo, balanced steady-state free precession, and single-shot rapid acquisition with relaxation enhancement. RT-MRI has earned an important role in both diagnostic imaging and image guidance of invasive procedures. Its unique diagnostic value is prominent in areas of the body that undergo substantial and often irregular motion, such as the heart, gastrointestinal system, upper airway vocal tract, and joints. Its value in interventional procedure guidance is prominent for procedures that require multiple forms of soft-tissue contrast, as well as flow information. In this review, we discuss the history of RT-MRI, fundamental tradeoffs, enabling technology, established applications, and current trends. LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY STAGE: 1.

Deep learning for radial SMS myocardial perfusion reconstruction using the 3D residual booster U-net

PURPOSE

To develop an end-to-end deep learning solution for quickly reconstructing radial simultaneous multi-slice (SMS) myocardial perfusion datasets with comparable quality to the pixel tracking spatiotemporal constrained reconstruction (PT-STCR) method.

METHODS

Dynamic contrast enhanced (DCE) radial SMS myocardial perfusion data were obtained from 20 subjects who were scanned at rest and/or stress with or without ECG gating using a saturation recovery radial CAIPI turboFLASH sequence. Input to the networks consisted of complex coil combined images reconstructed using the inverse Fourier transform of undersampled radial SMS k-space data. Ground truth images were reconstructed using the PT-STCR pipeline. The performance of the residual booster 3D U-Net was tested by comparing it to state-of-the-art network architectures including MoDL, CRNN-MRI, and other U-Net variants.

RESULTS

Results demonstrate significant improvements in speed requiring approximately 8 seconds to reconstruct one radial SMS dataset which is approximately 200 times faster than the PT-STCR method. Images reconstructed with the residual booster 3D U-Net retain quality of ground truth PT-STCR images (0.963 SSIM/40.238 PSNR/0.147 NRMSE). The residual booster 3D U-Net has superior performance compared to existing network architectures in terms of image quality, temporal dynamics, and reconstruction time.

CONCLUSION

Residual and booster learning combined with the 3D U-Net architecture was shown to be an effective network for reconstructing high-quality images from undersampled radial SMS datasets while bypassing the reconstruction time of the PT-STCR method.

Dynamic MRI of the abdomen using parallel non-Cartesian convolutional recurrent neural networks

PURPOSE

To improve the image quality and reduce computational time for the reconstruction of undersampled non-Cartesian abdominal dynamic parallel MR data using the deep learning approach.

METHODS

An algorithm of parallel non-Cartesian convolutional recurrent neural networks (PNCRNNs) was developed to enable the use of the redundant information in both spatial and temporal domains, and achieve data fidelity for the reconstruction of non-Cartesian parallel MR data. The performance of PNCRNNs was evaluated for various acceleration rates, motion patterns, and imaging applications in comparison with that of the state-of-the-art algorithms of dynamic imaging, including extra-dimensional golden-angle radial sparse parallel MRI (XD-GRASP), low-rank plus sparse matrix decomposition (L+S), blind compressive sensing (BCS), and 3D convolutional neural networks (3D CNNs).

RESULTS

PNCRNNs increased the peak SNR of 9.07 dB compared with XD-GRASP, 9.26 dB compared with L+S, 3.48 dB compared with BCS, and 3.14 dB compared with 3D CNN at R = 16. The reconstruction time was 18 ms for each bin, which was two orders faster than that of XD-GRASP, L+S, and BCS. PNCRNNs provided good reconstruction for various motion patterns, k-space trajectories, and imaging applications.

CONCLUSION

The proposed PNCRNN provides substantial improvement of the image quality for dynamic golden-angle radial imaging of the abdomen in comparison with XD-GRASP, L+S, BCS, and 3D CNN. The reconstruction time of PNCRNN can be as fast as 50 bins per second, due to the use of the highly computational efficient Toeplitz approach.

Closed-loop control of k-space sampling via physiologic feedback for cine MRI

BACKGROUND

Segmented cine cardiac MRI combines data from multiple heartbeats to achieve high spatiotemporal resolution cardiac images, yet predefined k-space segmentation trajectories can lead to suboptimal k-space sampling. In this work, we developed and evaluated an autonomous and closed-loop control system for radial k-space sampling (ARKS) to increase sampling uniformity.

METHODS

The closed-loop system autonomously selects radial k-space sampling trajectory during live segmented cine MRI and attempts to optimize angular sampling uniformity by selecting views in regions of k-space that were not previously well-sampled. Sampling uniformity and the ability to detect cardiac phase in vivo was assessed using ECG data acquired from 10 normal subjects in an MRI scanner. The approach was then implemented with a fast gradient echo sequence on a whole-body clinical MRI scanner and imaging was performed in 4 healthy volunteers. The closed-loop k-space trajectory was compared to random, uniformly distributed and golden angle view trajectories via measurement of k-space uniformity and the point spread function. Lastly, an arrhythmic dataset was used to evaluate a potential application of the approach.

RESULTS

The autonomous trajectory increased k-space sampling uniformity by 15±7%, main lobe point spread function (PSF) signal intensity by 6±4%, and reduced ringing relative to golden angle sampling. When implemented, the autonomous pulse sequence prescribed radial view angles faster than the scan TR (0.98 ± 0.01 ms, maximum = 1.38 ms) and increased k-space sampling mean uniformity by 10±11%, decreased uniformity variability by 44±12%, and increased PSF signal ratio by 6±6% relative to golden angle sampling.

CONCLUSION

The closed-loop approach enables near-uniform radial sampling in a segmented acquisition approach which was higher than predetermined golden-angle radial sampling. This can be utilized to increase the sampling or decrease the temporal footprint of an acquisition and the closed-loop framework has the potential to be applied to patients with complex heart rhythms.

Cardiac real-time MRI using a pre-emphasized spiral acquisition based on the gradient system transfer function

PURPOSE

Segmented Cartesian acquisition in breath hold represents the current gold standard for cardiac functional MRI. However, it is also associated with long imaging times and severe restrictions in arrhythmic or dyspneic patients. Therefore, we introduce a real-time imaging technique based on a spoiled gradient-echo sequence with undersampled spiral k-space trajectories corrected by a gradient pre-emphasis.

METHODS

A fully automatic gradient waveform pre-emphasis based on the gradient system transfer function was implemented to compensate for gradient inaccuracies, to optimize fast double-oblique spiral MRI. The framework was tested in a phantom study and subsequently transferred to compressed sensing-accelerated cardiac functional MRI in real time. Spiral acquisitions during breath hold and free breathing were compared with this reference method for healthy subjects (N = 7) as well as patients (N = 2) diagnosed with heart failure and arrhythmia. Left-ventricular volumes and ejection fractions were determined and analyzed using a Wilcoxon signed-rank test.

RESULTS

The pre-emphasis successfully reduced typical artifacts caused by k-space misregistrations. Dynamic cardiac imaging was possible in real time (temporal resolution < 50 ms) with high spatial resolution (1.34 × 1.34 mm² ), resulting in a total scan time of less than 50 seconds for whole heart coverage. Comparable image quality, as well as similar left-ventricular volumes and ejection fractions, were observed for the accelerated and the reference method.

CONCLUSION

The proposed technique enables high-resolution real-time cardiac MRI with no need for breath holds and electrocardiogram gating, shortening the duration of an entire functional cardiac exam to less than 1 minute.

Deep learning reconstruction for cardiac magnetic resonance fingerprinting T1 and T2 mapping

PURPOSE

To develop a deep learning method for rapidly reconstructing T₁ and T₂ maps from undersampled electrocardiogram (ECG) triggered cardiac magnetic resonance fingerprinting (cMRF) images.

METHODS

A neural network was developed that outputs T₁ and T₂ values when given a measured cMRF signal time course and cardiac RR interval times recorded by an ECG. Over 8 million cMRF signals, corresponding to 4000 random cardiac rhythms, were simulated for training. The training signals were corrupted by simulated k-space undersampling artifacts and random phase shifts to promote robust learning. The deep learning reconstruction was evaluated in Monte Carlo simulations for a variety of cardiac rhythms and compared with dictionary-based pattern matching in 58 healthy subjects at 1.5T.

RESULTS

In simulations, the normalized root-mean-square error (nRMSE) for T₁ was below 1% in myocardium, blood, and liver for all tested heart rates. For T₂ , the nRMSE was below 4% for myocardium and liver and below 6% for blood for all heart rates. The difference in the mean myocardial T₁ or T₂ observed in vivo between dictionary matching and deep learning was 3.6 ms for T₁ and -0.2 ms for T₂ . Whereas dictionary generation and pattern matching required more than 4 min per slice, the deep learning reconstruction only required 336 ms.

CONCLUSION

A neural network is introduced for reconstructing cMRF T₁ and T₂ maps directly from undersampled spiral images in under 400 ms and is robust to arbitrary cardiac rhythms, which paves the way for rapid online display of cMRF maps.

Cardiac and Respiratory Self-Gating in Radial MRI Using an Adapted Singular Spectrum Analysis (SSA-FARY)

Cardiac Magnetic Resonance Imaging (MRI) is time-consuming and error-prone. To ease the patient's burden and to increase the efficiency and robustness of cardiac exams, interest in methods based on continuous steady-state acquisition and self-gating has been growing in recent years. Self-gating methods extract the cardiac and respiratory signals from the measurement data and then retrospectively sort the data into cardiac and respiratory phases. Repeated breathholds and synchronization with the heart beat using some external device as required in conventional MRI are then not necessary. In this work, we introduce a novel self-gating method for radially acquired data based on a dimensionality reduction technique for time-series analysis (SSA-FARY). Building on Singular Spectrum Analysis, a zero-padded, time-delayed embedding of the auto-calibration data is analyzed using Principle Component Analysis. We demonstrate the basic functionality of SSA-FARY using numerical simulations and apply it to in-vivo cardiac radial single-slice bSSFP and Simultaneous Multi-Slice radiofrequency-spoiled gradient-echo measurements, as well as to Stack-of-Stars bSSFP measurements. SSA-FARY reliably detects the cardiac and respiratory motion and separates it from noise. We utilize the generated signals for high-dimensional image reconstruction using parallel imaging and compressed sensing with in-plane wavelet and (spatio-)temporal total-variation regularization.

Minimal Linear Networks for Magnetic Resonance Image Reconstruction

Modern sequences for Magnetic Resonance Imaging (MRI) trade off scan time with computational challenges, resulting in ill-posed inverse problems and the requirement to account for more elaborated signal models. Various deep learning techniques have shown potential for image reconstruction from reduced data, outperforming compressed sensing, dictionary learning and other advanced techniques based on regularization, by characterization of the image manifold. In this work we suggest a framework for reducing a “neural” network to the bare minimum required by the MR physics, reducing the network depth and removing all non-linearities. The networks performed well both on benchmark simulated data and on arterial spin labeling perfusion imaging, showing clear images while preserving sensitivity to the minute signal changes. The results indicate that the deep learning framework plays a major role in MR image reconstruction, and suggest a concrete approach for probing into the contribution of additional elements.

Improvement of Fast Model-Based Acceleration of Parameter Look-Locker T1 Mapping

Quantitative mapping is desirable in many scientific and clinical magneric resonance imaging (MRI) applications. Recent inverse recovery-look locker sequence enables single-shot T₁ mapping with a time of a few seconds but the main computational load is directed into offline reconstruction, which can take from several minutes up to few hours. In this study we proposed improvement of model-based approach for T₁-mapping by introduction of two steps fitting procedure. We provided analysis of further reduction of k-space data, which lead us to decrease of computational time and perform simulation of multi-slice development. The region of interest (ROI) analysis of human brain measurements with two different initial models shows that the differences between mean values with respect to a reference approach are in white matter-0.3% and 1.1%, grey matter-0.4% and 1.78% and cerebrospinal fluid-2.8% and 11.1% respectively. With further improvements we were able to decrease the time of computational of single slice to 6.5 min and 23.5 min for different initial models, which has been already not achieved by any other algorithm. In result we obtained an accelerated novel method of model-based image reconstruction in which single iteration can be performed within few seconds on home computer.

Self-calibrated interpolation of non-Cartesian data with GRAPPA in parallel imaging

PURPOSE

To develop a non-Cartesian k-space reconstruction method using self-calibrated region-specific interpolation kernels for highly accelerated acquisitions.

METHODS

In conventional non-Cartesian GRAPPA with through-time GRAPPA (TT-GRAPPA), the use of region-specific interpolation kernels has demonstrated improved reconstruction quality in dynamic imaging for highly accelerated acquisitions. However, TT-GRAPPA requires the acquisition of a large number of separate calibration scans. To reduce the overall imaging time, we propose Self-calibrated Interpolation of Non-Cartesian data with GRAPPA (SING) to self-calibrate region-specific interpolation kernels from dynamic undersampled measurements. The SING method synthesizes calibration data to adapt to the distinct shape of each region-specific interpolation kernel geometry, and uses a novel local k-space regularization through an extension of TT-GRAPPA. This calibration approach is used to reconstruct non-Cartesian images at high acceleration rates while mitigating noise amplification. The reconstruction quality of SING is compared with conjugate-gradient SENSE and TT-GRAPPA in numerical phantoms and in vivo cine data sets.

RESULTS

In both numerical phantom and in vivo cine data sets, SING offers visually and quantitatively similar reconstruction quality to TT-GRAPPA, and provides improved reconstruction quality over conjugate-gradient SENSE. Furthermore, temporal fidelity in SING and TT-GRAPPA is similar for the same acceleration rates. G-factor evaluation over the heart shows that SING and TT-GRAPPA provide similar noise amplification at moderate and high rates.

CONCLUSION

The proposed SING reconstruction enables significant improvement of acquisition efficiency for calibration data, while matching the reconstruction performance of TT-GRAPPA.

Short-T2 MRI: Principles and recent advances

Among current modalities of biomedical and diagnostic imaging, MRI stands out by virtue of its versatile contrast obtained without ionizing radiation. However, in various cases, e.g., water protons in tissues such as bone, tendon, and lung, MRI performance is limited by the rapid decay of resonance signals associated with short transverse relaxation times T₂ or T₂*. Efforts to address this shortcoming have led to a variety of specialized short-T₂ techniques. Recent progress in this field expands the choice of methods and prompts fresh considerations with regard to instrumentation, data acquisition, and signal processing. In this review, the current status of short-T₂ MRI is surveyed. In an attempt to structure the growing range of techniques, the presentation highlights overarching concepts and basic methodological options. The most frequently used approaches are described in detail, including acquisition strategies, image reconstruction, hardware requirements, means of introducing contrast, sources of artifacts, limitations, and applications.

Orthogonal tensor dictionary learning for accelerated dynamic MRI

A direct application of the compressed sensing (CS) theory to dynamic magnetic resonance imaging (MRI) reconstruction needs vectorization or matricization of the dynamic MRI data, which is composed of a stack of 2D images and can be naturally regarded as a tensor. This 1D/2D model may destroy the inherent spatial structure property of the data. An alternative way to exploit the multidimensional structure in dynamic MRI is to employ tensor decomposition for dictionary learning, that is, learning multiple dictionaries along each dimension (mode) and sparsely representing the multidimensional data with respect to the Kronecker product of these dictionaries. In this work, we introduce a novel tensor dictionary learning method under an orthonormal constraint on the elementary matrix of the tensor dictionary for dynamic MRI reconstruction. The proposed algorithm alternates sparse coding, tensor dictionary learning, and updating reconstruction, and each corresponding subproblem is efficiently solved by a closed-form solution. Numerical experiments on phantom and synthetic data show significant improvements in reconstruction accuracy and computational efficiency obtained by the proposed scheme over the existing method that uses the 1D/2D model with overcomplete dictionary learning. Graphical abstract Fig. 1 Comparison between (a) the traditional method and (b) the proposed method based on dictionary learning for dynamic MRI reconstruction.

A GRAPPA algorithm for arbitrary 2D/3D non-Cartesian sampling trajectories with rapid calibration

PURPOSE

GRAPPA is a popular reconstruction method for Cartesian parallel imaging, but is not easily extended to non-Cartesian sampling. We introduce a general and practical GRAPPA algorithm for arbitrary non-Cartesian imaging.

METHODS

We formulate a general GRAPPA reconstruction by associating a unique kernel with each unsampled k-space location with a distinct constellation, that is, local sampling pattern. We calibrate these generalized kernels using the Fourier transform phase shift property applied to fully gridded or separately acquired Cartesian Autocalibration signal (ACS) data. To handle the resulting large number of different kernels, we introduce a fast calibration algorithm based on nonuniform FFT (NUFFT) and adoption of circulant ACS boundary conditions. We applied our method to retrospectively under-sampled rotated stack-of-stars/spirals in vivo datasets, and to a prospectively under-sampled rotated stack-of-spirals functional MRI acquisition with a finger-tapping task.

RESULTS

We reconstructed all datasets without performing any trajectory-specific manual adaptation of the method. For the retrospectively under-sampled experiments, our method achieved image quality (i.e., error and g-factor maps) comparable to conjugate gradient SENSE (cg-SENSE) and SPIRiT. Functional activation maps obtained from our method were in good agreement with those obtained using cg-SENSE, but required a shorter total reconstruction time (for the whole time-series): 3 minutes (proposed) vs 15 minutes (cg-SENSE).

CONCLUSIONS

This paper introduces a general 3D non-Cartesian GRAPPA that is fast enough for practical use on today's computers. It is a direct generalization of original GRAPPA to non-Cartesian scenarios. The method should be particularly useful in dynamic imaging where a large number of frames are reconstructed from a single set of ACS data.

Partial fourier shells trajectory for non-cartesian MRI

Non-Cartesian MRI acquisition has demonstrated various advantages in many clinical applications. The shells trajectory is a 3D non-Cartesian MRI acquisition technique that samples the k-space using a series of concentric shells to achieve efficient 3D isotropic acquisition. Partial Fourier acquisition is an acceleration technique that is widely used in Cartesian MRI. It exploits the conjugate symmetry of k-space measurement to reduce the number of k-space samples compared to full-k-space acquisition, without loss of spatial resolution. For a Cartesian MRI acquisition, the direction of partial Fourier acceleration is aligned either with the phase encoded or frequency encoded direction. In those cases, the underlying image matrix can be reconstructed from the undersampled k-space data using a non-iterative, homodyne reconstruction framework. However, designing a non-Cartesian acquisition trajectory that is compatible with non-iterative homodyne reconstruction is not nearly as straightforward as in the Cartesian case. One reason is the non-iterative homodyne reconstruction requires (slightly over) half of the k-space to be fully sampled. Since the direction of partial Fourier acceleration varies throughout the acquisition in the non-Cartesian trajectory, directly applying the same partial Fourier acquisition pattern (as in Cartesian acquisitions) to a non-Cartesian trajectory does not necessarily yield a continuous, physically-achievable trajectory. In this work, we develop an asymmetric shells trajectory with fully-automated trajectory and gradient waveform design to achieve partial Fourier acquisition for the shells trajectory. We then demonstrate a non-iterative image reconstruction framework for the proposed trajectory. Phantom and in vivo brain scans based on spoiled gradient echo (SPGR) shells and magnetization-prepared shells (MP-shells) were performed to test the proposed trajectory design and reconstruction method. Our phantom and in vivo results demonstrate that the proposed partial Fourier shells trajectory maintains the desirable image contrast and high sampling efficiency from the fully sampled shells, while further reducing data acquisition time.

Simultaneous T1 and T2 measurements using inversion recovery TrueFISP with principle component-based reconstruction, off-resonance correction, and multicomponent analysis

PURPOSE

To improve the reconstruction quality for quantitative T₁ and T₂ measurements using the inversion recovery (IR) TrueFISP sequence and to demonstrate the potential for multicomponent analysis.

METHODS

The iterative reconstruction method takes advantage of the high redundancy in the smooth exponential signals using principle component analysis (PCA). Multicomponent information is preserved and allows voxel-by-voxel computation of relaxation time spectra with an inverse Laplace transform. Off-resonance effects are analytically and numerically investigated and a correction approach is presented.

RESULTS

Single-shot IR TrueFISP in vivo measurements on healthy volunteers demonstrate the improved reconstruction performance compared to a view sharing (k-space weighted image contrast [KWIC]) reconstruction. Especially, tissue components with short apparent relaxation times T₁ ^* are not filtered out and can be identified in the relaxation time spectra. These components include myelin in the human brain (T₁ ^* ≈ 130 ms) and extra cranial subcutaneous fat.

CONCLUSION

The PCA-based reconstruction method improves the temporal accuracy and preserves multicomponent information. Spatially resolved relaxation time spectra can be obtained and allow the identification of tissue types with short, apparent relaxation times.

Real-time cardiac MRI with radial acquisition and k-space variant reduced-FOV reconstruction

This work aims to demonstrate that radial acquisition with k-space variant reduced-FOV reconstruction can enable real-time cardiac MRI with an affordable computation cost. Due to non-uniform sampling, radial imaging requires k-space variant reconstruction for optimal performance. By converting radial parallel imaging reconstruction into the estimation of correlation functions with a previously-developed correlation imaging framework, Cartesian k-space may be reconstructed point-wisely based on parallel imaging relationship between every Cartesian datum and its neighboring radial samples. Furthermore, reduced-FOV correlation functions may be used to calculate a subset of Cartesian k-space data for image reconstruction within a small region of interest, making it possible to run real-time cardiac MRI with an affordable computation cost. In a stress cardiac test where the subject is imaged during biking with a heart rate of >100 bpm, this k-space variant reduced-FOV reconstruction is demonstrated in reference to several radial imaging techniques including gridding, GROG and SPIRiT. It is found that the k-space variant reconstruction outperforms gridding, GROG and SPIRiT in real-time imaging. The computation cost of reduced-FOV reconstruction is ~2 times higher than that of GROG. The presented work provides a practical solution to real-time cardiac MRI with radial acquisition and k-space variant reduced-FOV reconstruction in clinical settings.

RACER-GRASP: Respiratory-weighted, aortic contrast enhancement-guided and coil-unstreaking golden-angle radial sparse MRI

PURPOSE

To develop and evaluate a novel dynamic contrast-enhanced imaging technique called RACER-GRASP (Respiratory-weighted, Aortic Contrast Enhancement-guided and coil-unstReaking Golden-angle RAdial Sparse Parallel) MRI that extends GRASP to include automatic contrast bolus timing, respiratory motion compensation, and coil-weighted unstreaking for improved imaging performance in liver MRI.

METHODS

In RACER-GRASP, aortic contrast enhancement (ACE) guided k-space sorting and respiratory-weighted sparse reconstruction are performed using aortic contrast enhancement and respiratory motion signals extracted directly from the acquired data. Coil unstreaking aims to weight multicoil k-space according to streaking artifact level calculated for each individual coil during image reconstruction, so that coil elements containing a high level of streaking artifacts contribute less to the final results. Self-calibrating GRAPPA operator gridding was applied as a pre-reconstruction step to reduce computational burden in the subsequent iterative reconstruction. The RACER-GRASP technique was compared with standard GRASP reconstruction in a group of healthy volunteers and patients referred for clinical liver MR examination.

RESULTS

Compared with standard GRASP, RACER-GRASP significantly improved overall image quality (average score: 3.25 versus 3.85) and hepatic vessel sharpness/clarity (average score: 3.58 versus 4.0), and reduced residual streaking artifact level (average score: 3.23 versus 3.94) in different contrast phases. RACER-GRASP also enabled automatic timing of the arterial phases.

CONCLUSIONS

The aortic contrast enhancement-guided sorting, respiratory motion suppression and coil unstreaking introduced by RACER-GRASP improve upon the imaging performance of standard GRASP for free-breathing dynamic contrast-enhanced MRI of the liver. Magn Reson Med 80:77-89, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

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.

3D MR fingerprinting with accelerated stack-of-spirals and hybrid sliding-window and GRAPPA reconstruction

PURPOSE

Whole-brain high-resolution quantitative imaging is extremely encoding intensive, and its rapid and robust acquisition remains a challenge. Here we present a 3D MR fingerprinting (MRF) acquisition with a hybrid sliding-window (SW) and GRAPPA reconstruction strategy to obtain high-resolution T1, T2 and proton density (PD) maps with whole brain coverage in a clinically feasible timeframe.

METHODS

3D MRF data were acquired using a highly under-sampled stack-of-spirals trajectory with a steady-state precession (FISP) sequence. For data reconstruction, kx-ky under-sampling was mitigated using SW combination along the temporal axis. Non-uniform fast Fourier transform (NUFFT) was then applied to create Cartesian k-space data that are fully-sampled in the in-plane direction, and Cartesian GRAPPA was performed to resolve kz under-sampling to create an alias-free SW dataset. T1, T2 and PD maps were then obtained using dictionary matching.

RESULTS

Phantom study demonstrated that the proposed 3D-MRF acquisition/reconstruction method is able to produce quantitative maps that are consistent with conventional quantification techniques. Retrospectively under-sampled in vivo acquisition revealed that SW + GRAPPA substantially improves quantification accuracy over the current state-of-the-art accelerated 3D MRF. Prospectively under-sampled in vivo study showed that whole brain T1, T2 and PD maps with 1 mm3 resolution could be obtained in 7.5 min.

CONCLUSIONS

3D MRF stack-of-spirals acquisition with hybrid SW + GRAPPA reconstruction may provide a feasible approach for rapid, high-resolution quantitative whole-brain imaging.

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.

Use of pattern recognition for unaliasing simultaneously acquired slices in simultaneous multislice MR fingerprinting

PURPOSE

The purpose of this study is to accelerate an MR fingerprinting (MRF) acquisition by using a simultaneous multislice method.

METHODS

A multiband radiofrequency (RF) pulse was designed to excite two slices with different flip angles and phases. The signals of two slices were driven to be as orthogonal as possible. The mixed and undersampled MRF signal was matched to two dictionaries to retrieve T₁ and T₂ maps of each slice. Quantitative results from the proposed method were validated with the gold-standard spin echo methods in a phantom. T₁ and T₂ maps of in vivo human brain from two simultaneously acquired slices were also compared to the results of fast imaging with steady-state precession based MRF method (MRF-FISP) with a single-band RF excitation.

RESULTS

The phantom results showed that the simultaneous multislice imaging MRF-FISP method quantified the relaxation properties accurately compared to the gold-standard spin echo methods. T₁ and T₂ values of in vivo brain from the proposed method also matched the results from the normal MRF-FISP acquisition.

CONCLUSION

T₁ and T₂ values can be quantified at a multiband acceleration factor of two using our proposed acquisition even in a single-channel receive coil. Further acceleration could be achieved by combining this method with parallel imaging or iterative reconstruction. Magn Reson Med 78:1870-1876, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

MR fingerprinting using the quick echo splitting NMR imaging technique

PURPOSE

The purpose of the study is to develop a quantitative method for the relaxation properties with a reduced radio frequency (RF) power deposition by combining magnetic resonance fingerprinting (MRF) technique with quick echo splitting NMR imaging technique (QUEST).

METHODS

A QUEST-based MRF sequence was implemented to acquire high-order echoes by increasing the gaps between RF pulses. Bloch simulations were used to calculate a dictionary containing the range of physically plausible signal evolutions using a range of T₁ and T₂ values based on the pulse sequence. MRF-QUEST was evaluated by comparing to the results of spin-echo methods. The specific absorption rate (SAR) of MRF-QUEST was compared with the clinically available methods.

RESULTS

MRF-QUEST quantifies the relaxation properties with good accuracy at the estimated head SAR of 0.03 W/kg. T₁ and T₂ values estimated by MRF-QUEST are in good agreement with the traditional methods.

CONCLUSIONS

The combination of the MRF and the QUEST provides an accurate quantification of T₁ and T₂ simultaneously with reduced RF power deposition. The resulting lower SAR may provide a new acquisition strategy for MRF when RF energy deposition is problematic. Magn Reson Med 77:979-988, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

A model-based reconstruction technique for fast dynamic T1 mapping

PURPOSE

To present a technique for dynamic T1 mapping.

MATERIALS AND METHODS

A recently proposed model-based reconstruction entitled IR-MAP allows T1 mapping of a single slice from a single radial inversion recovery Look-Locker FLASH acquisition. To enable dynamic T1 mapping, multiple of these acquisitions are consecutively performed, each followed by a waiting period of 3s for relaxation. Next, IR-MAP is used to reconstruct an individual T1 map for each of these acquisitions. Finally, T1 errors caused by insufficient relaxation between subsequent IR pulses are iteratively corrected.

RESULTS

The functionality of the proposed setup was validated in a phantom and in seven healthy volunteers. Systematic deviations between subsequent T1 maps originating from insufficient relaxation periods were effectively corrected. Additionally, the approach was successfully applied to monitor the T1 dynamic in a patient with primary lymphoma after the intravenous injection of contrast agent.

CONCLUSION

The proposed setup enables dynamic T1 mapping of a single slice with a spatial resolution of 1.6 mm × 1.6 mm × 3 mm and a temporal resolution of one parameter map every 9 s. It therefore represents a new opportunity to track changes in T1 over time, as it is desirable in many applications such as dynamic contrast-enhanced MRI.

Evaluation of optimized breath-hold and free-breathing 3D ultrashort echo time contrast agent-free MRI of the human lung

PURPOSE

To evaluate an optimized stack of radials ultrashort echo time (UTE) 3D magnetic resonance imaging (MRI) sequence for breath-hold and free-breathing imaging of the human lung.

MATERIALS AND METHODS

A 3D stack of ultrashort echo time radials trajectory was optimized for coronal and axial lower-resolution breath-hold and higher-resolution free-breathing scans using Bloch simulations. The sequence was evaluated in 10 volunteers, without the use of contrast agents. Signal-to-noise ratio (SNR) mean and 95% confidence interval (CI) were determined from separate signal and noise images in a semiautomated fashion. The four scanning schemes were evaluated for significant differences in image quality using Student's t-test. Ten clinical patients were scanned with the sequence and findings were compared with concomitant computed tomography (CT) in nine patients. Breath-hold 3D spokes images were compared with 3D stack of radials in five volunteers. A Mann-Whitney U-test was performed to test significance in both cases.

RESULTS

Breath-hold imaging of the entire lung in volunteers was performed with SNR (mean = 42.5 [CI]: 35.5-49.5; mean = 34.3 [CI]: 28.6-40) in lung parenchyma for coronal and axial scans, respectively, which can be used as a quick scout scan. Longer respiratory triggered free-breathing scan enabled high-resolution UTE scanning with mean SNR of 14.2 ([CI]: 12.9-15.5) and 9.2 ([CI]: 8.2-10.2) for coronal and axial scans, respectively. Axial free-breathing scans showed significantly higher image quality (P = 0.008) than the three other scanning schemes. The mean score for comparison with CT was 1.67 (score 0: n = 0; 1: n = 3; 2: n = 6). There was no significant difference between CT and MRI (P = 0.25). 3D stack of radials images were significantly better than 3D spokes images (P < 0.001).

CONCLUSION

The optimized 3D stack of radials trajectory was shown to provide high-quality MR images of the lung parenchyma without the use of MRI contrast agents. The sequence may offer the possibility of breath-hold imaging and provides greater flexibility in trading off slice thickness and parallel imaging for scan time.

Reducing acquisition time in clinical MRI by data undersampling and compressed sensing reconstruction

MRI is often the most sensitive or appropriate technique for important measurements in clinical diagnosis and research, but lengthy acquisition times limit its use due to cost and considerations of patient comfort and compliance. Once an image field of view and resolution is chosen, the minimum scan acquisition time is normally fixed by the amount of raw data that must be acquired to meet the Nyquist criteria. Recently, there has been research interest in using the theory of compressed sensing (CS) in MR imaging to reduce scan acquisition times. The theory argues that if our target MR image is sparse, having signal information in only a small proportion of pixels (like an angiogram), or if the image can be mathematically transformed to be sparse then it is possible to use that sparsity to recover a high definition image from substantially less acquired data. This review starts by considering methods of k-space undersampling which have already been incorporated into routine clinical imaging (partial Fourier imaging and parallel imaging), and then explains the basis of using compressed sensing in MRI. The practical considerations of applying CS to MRI acquisitions are discussed, such as designing k-space undersampling schemes, optimizing adjustable parameters in reconstructions and exploiting the power of combined compressed sensing and parallel imaging (CS-PI). A selection of clinical applications that have used CS and CS-PI prospectively are considered. The review concludes by signposting other imaging acceleration techniques under present development before concluding with a consideration of the potential impact and obstacles to bringing compressed sensing into routine use in clinical MRI.

A model-based reconstruction technique for quantitative myocardial perfusion imaging

PURPOSE

To reduce saturation effects in the arterial input function (AIF) estimation of quantitative myocardial first-pass saturation recovery perfusion imaging by employing a model-based reconstruction.

THEORY AND METHODS

Imaging was performed with a saturation recovery prepared radial FLASH sequence. A model-based reconstruction was applied for reconstruction. By exploiting prior knowledge about the relaxation process, an image series with different saturation recovery times was reconstructed. By evaluating images with an effective saturation time of approximately 3 ms, saturation effects in the AIF determination were reduced. In a volunteer study, this approach was compared with a standard prebolus technique.

RESULTS

In comparison to the low-dose injection of a prebolus acquisition, saturation effects were further reduced in the AIFs determined using the model-based approach. These effects, which were clearly visible for all six volunteers, were reflected in a statistically significant difference of up to 20% in the absolute perfusion values.

CONCLUSION

The application of model-based reconstruction algorithms in quantitative myocardial perfusion imaging promises a significant improvement of the AIF determination. In addition to greatly reducing saturation effects that occur even for the prebolus methods, only a single bolus has to be applied. Magn Reson Med 76:880-887, 2016. © 2015 Wiley Periodicals, Inc.

Trajectory optimization based on the signal-to-noise ratio for spatial encoding with nonlinear encoding fields

PURPOSE

Multiple nonlinear gradient fields offer many potential benefits for spatial encoding including reduced acquisition time, fewer artefacts and region-specific imaging, although designing a suitable trajectory for such a setup is difficult. This work aims to optimize encoding trajectories for multiple nonlinear gradient fields based on the image signal-to-noise ratio.

THEORY AND METHODS

Image signal-to-noise ratio is directly linked to the covariance of the reconstructed pixels, which can be calculated recursively for each projection of the trajectory under a Bayesian formulation. An evolutionary algorithm is used to find the higher-dimensional projections that minimize the pixel covariance, incorporating receive coil profiles, intravoxel dephasing, and reconstruction regularization. The resulting trajectories are tested through simulations and experiments.

RESULTS

The optimized trajectories produce images with higher resolution and fewer artefacts compared with traditional approaches, particularly for high undersampling. However, higher-dimensional projection experiments strongly depend on accurate hardware and calibration.

CONCLUSION

Computer-based optimization provides an efficient means to explore the large trajectory space created by the use of multiple nonlinear encoding fields. The optimization framework, as presented here, is necessary to fully exploit the advantages of nonlinear fields. Magn Reson Med 76:104-117, 2016. © 2015 Wiley Periodicals, Inc.

Accelerating magnetic resonance fingerprinting (MRF) using t-blipped simultaneous multislice (SMS) acquisition

PURPOSE

We incorporate simultaneous multislice (SMS) acquisition into MR fingerprinting (MRF) to accelerate the MRF acquisition.

METHODS

The t-Blipped SMS-MRF method is achieved by adding a Gz blip before each data acquisition window and balancing it with a Gz blip of opposing polarity at the end of each TR. Thus the signal from different simultaneously excited slices are encoded with different phases without disturbing the signal evolution. Furthermore, by varying the Gz blip area and/or polarity as a function of repetition time, the slices' differential phase can also be made to vary as a function of time. For reconstruction of t-Blipped SMS-MRF data, we demonstrate a combined slice-direction SENSE and modified dictionary matching method.

RESULTS

In Monte Carlo simulation, the parameter mapping from multiband factor (MB) = 2 t-Blipped SMS-MRF shows good accuracy and precision when compared with results from reference conventional MRF data with concordance correlation coefficients (CCC) of 0.96 for T1 estimates and 0.90 for T2 estimates. For in vivo experiments, T1 and T2 maps from MB=2 t-Blipped SMS-MRF have a high agreement with ones from conventional MRF.

CONCLUSION

The MB=2 t-Blipped SMS-MRF acquisition/reconstruction method has been demonstrated and validated to provide more rapid parameter mapping in the MRF framework.

Model-based acceleration of look-locker T1 mapping

Mapping the longitudinal relaxation time T₁ has widespread applications in clinical MRI as it promises a quantitative comparison of tissue properties across subjects and scanners. Due to the long scan times of conventional methods, however, the use of quantitative MRI in clinical routine is still very limited. In this work, an acceleration of Inversion-Recovery Look-Locker (IR-LL) T₁ mapping is presented. A model-based algorithm is used to iteratively enforce an exponential relaxation model to a highly undersampled radially acquired IR-LL dataset obtained after the application of a single global inversion pulse. Using the proposed technique, a T₁ map of a single slice with 1.6mm in-plane resolution and 4mm slice thickness can be reconstructed from data acquired in only 6s. A time-consuming segmented IR experiment was used as gold standard for T₁ mapping in this work. In the subsequent validation study, the model-based reconstruction of a single-inversion IR-LL dataset exhibited a T₁ difference of less than 2.6% compared to the segmented IR-LL reference in a phantom consisting of vials with T₁ values between 200ms and 3000ms. In vivo, the T₁ difference was smaller than 5.5% in WM and GM of seven healthy volunteers. Additionally, the T₁ values are comparable to standard literature values. Despite the high acceleration, all model-based reconstructions were of a visual quality comparable to fully sampled references. Finally, the reproducibility of the T₁ mapping method was demonstrated in repeated acquisitions. In conclusion, the presented approach represents a promising way for fast and accurate T₁ mapping using radial IR-LL acquisitions without the need of any segmentation.

State-of-the-art MRI techniques in neuroradiology: principles, pitfalls, and clinical applications

This article reviews the most relevant state-of-the-art magnetic resonance (MR) techniques, which are clinically available to investigate brain diseases. MR acquisition techniques addressed include notably diffusion imaging (diffusion-weighted imaging (DWI), diffusion tensor imaging (DTI), and diffusion kurtosis imaging (DKI)) as well as perfusion imaging (dynamic susceptibility contrast (DSC), arterial spin labeling (ASL), and dynamic contrast enhanced (DCE)). The underlying models used to process these images are described, as well as the theoretic underpinnings of quantitative diffusion and perfusion MR imaging-based methods. The technical requirements and how they may help to understand, classify, or follow-up neurological pathologies are briefly summarized. Techniques, principles, advantages but also intrinsic limitations, typical artifacts, and alternative solutions developed to overcome them are discussed. In this article, we also review routinely available three-dimensional (3D) techniques in neuro MRI, including state-of-the-art and emerging angiography sequences, and briefly introduce more recently proposed 3D quantitative neuro-anatomy sequences, and new technology, such as multi-slice and multi-transmit imaging.

Parallel Magnetic Resonance Imaging as Approximation in a Reproducing Kernel Hilbert Space

In Magnetic Resonance Imaging (MRI) data samples are collected in the spatial frequency domain (k-space), typically by time-consuming line-by-line scanning on a Cartesian grid. Scans can be accelerated by simultaneous acquisition of data using multiple receivers (parallel imaging), and by using more efficient non-Cartesian sampling schemes. To understand and design k-space sampling patterns, a theoretical framework is needed to analyze how well arbitrary sampling patterns reconstruct unsampled k-space using receive coil information. As shown here, reconstruction from samples at arbitrary locations can be understood as approximation of vector-valued functions from the acquired samples and formulated using a Reproducing Kernel Hilbert Space (RKHS) with a matrix-valued kernel defined by the spatial sensitivities of the receive coils. This establishes a formal connection between approximation theory and parallel imaging. Theoretical tools from approximation theory can then be used to understand reconstruction in k-space and to extend the analysis of the effects of samples selection beyond the traditional image-domain g-factor noise analysis to both noise amplification and approximation errors in k-space. This is demonstrated with numerical examples.

Self-calibrated trajectory estimation and signal correction method for robust radial imaging using GRAPPA operator gridding

PURPOSE

In radial imaging, projections may become "miscentered" due to gradient errors such as delays and eddy currents. These errors may result in image artifacts and can disrupt the reliability of direct current (DC) navigation. The proposed parallel imaging-based technique retrospectively estimates trajectory error from miscentered radial data without extra acquisitions, hardware, or sequence modification.

THEORY AND METHODS

After phase correction, self-calibrated GRAPPA operator gridding (GROG) weights are iteratively applied to shift-miscentered projections toward the center of k-space. A search algorithm identifies the shift that aligns the peak k-space signals by maximizing the sum-of-squares DC signal estimate of each projection. The algorithm returns a trajectory estimate and a corrected radial k-space signal.

RESULTS

Data from a spherical phantom, the head, and the heart demonstrate that image reconstruction with the estimated trajectory restores image quality and reduces artifacts such as streaks and signal voids. The DC signal level is increased and variability is reduced.

CONCLUSION

Retrospective phase correction and iterative application of GROG can be used to successfully estimate the trajectory error in two-dimensional radial acquisitions for improved image reconstruction without requiring extra data acquisition or sequence modification.

An augmented Lagrangian based compressed sensing reconstruction for non-Cartesian magnetic resonance imaging without gridding and regridding at every iteration

Background

Non-Cartesian trajectories are used in a variety of fast imaging applications, due to the incoherent image domain artifacts they create when undersampled. While the gridding technique is commonly utilized for reconstruction, the incoherent artifacts may be further removed using compressed sensing (CS). CS reconstruction is typically done using conjugate-gradient (CG) type algorithms, which require gridding and regridding to be performed at every iteration. This leads to a large computational overhead that hinders its applicability.

Methods

We sought to develop an alternative method for CS reconstruction that only requires two gridding and one regridding operation in total, irrespective of the number of iterations. This proposed technique is evaluated on phantom images and whole-heart coronary MRI acquired using 3D radial trajectories, and compared to conventional CS reconstruction using CG algorithms in terms of quantitative vessel sharpness, vessel length, computation time, and convergence rate.

Results

Both CS reconstructions result in similar vessel length (P = 0.30) and vessel sharpness (P = 0.62). The per-iteration complexity of the proposed technique is approximately 3-fold lower than the conventional CS reconstruction (17.55 vs. 52.48 seconds in C++). Furthermore, for in-vivo datasets, the convergence rate of the proposed technique is faster (60±13 vs. 455±320 iterations) leading to a ∼23-fold reduction in reconstruction time.

Conclusions

The proposed reconstruction provides images of similar quality to the conventional CS technique in terms of removing artifacts, but at a much lower computational complexity.

Consideration of slice profiles in inversion recovery Look-Locker relaxation parameter mapping

PURPOSE

To include the flip angle distribution caused by the slice profile into the model used for describing the relaxation curves observed in inversion recovery Look-Locker FLASH T1 mapping for a more accurate determination of the relaxation parameters.

MATERIALS AND METHODS

For each inversion time, the flip angle dependent signal of the mono-exponential relaxation model is integrated across the slice profile. The resulting Consideration of Slice Profiles (CSP) relaxation curves are compared to the mono-exponential signal model in numerical simulations as well as in phantom and in-vivo experiments.

RESULTS

All measured relaxation curves showed systematic deviations from a mono-exponential curve increasing with flip angle and T1 but decreasing with repetition time. Additionally, the accuracy of T1 was found to be largely dependent on the temporal coverage of the relaxation curve. All these systematic errors were largely reduced by the CSP model.

CONCLUSION

The proposed CSP model represents a useful extension of the conventionally used mono-exponential relaxation model. Despite inherent model inaccuracies, the mono-exponential model was found to be sufficient for many T1 mapping situations. However, if only a poor temporal coverage of the relaxation process is achievable or a very precise modeling of the relaxation course is needed as in model-based techniques, the mono-exponential model leads to systematic errors and the CSP model should be used instead.

Correlation imaging with arbitrary sampling trajectories

The presented work aims to develop a generalized linear approach to image reconstruction with arbitrary sampling trajectories for high-speed MRI. This approach is based on a previously developed image reconstruction framework, "correlation imaging". In the presented work, correlation imaging with arbitrary sampling trajectories is implemented in a multi-dimensional hybrid space that is formed from the physical sampling space and a virtually defined space. By introducing an undersampling trajectory with both uniformity and randomness in the hybrid space, correlation imaging may take advantage of multiple image reconstruction mechanisms including coil sensitivity encoding, data sparsity and information sharing. This hybrid-space implementation is demonstrated in multi-slice 2D imaging, multi-scan imaging, and radial dynamic imaging. Since more information is used in image reconstruction, it is found that hybrid-space correlation imaging outperforms several conventional techniques. The presented approach will benefit clinical MRI by enabling correlation imaging to be used to accelerate multi-scan clinical protocols that need different sampling trajectories in different scans.

Using self-consistency for an iterative trajectory adjustment (SCITA)

PURPOSE

To iteratively correct for deviations in radial trajectories with no need of additionally performed calibration scans.

THEORY AND METHODS

Radially acquired data sets-even when undersampled to a certain extend-inherently feature an oversampled area in the center of k-space. Thus, for a perfectly measured trajectory and neglecting noise, information is consistent between multiple measurements gridded to the same Cartesian position within this region. In the case of erroneous coordinates, this accordance-and therefore a correction of the trajectory-can be enforced by an algorithm iteratively shifting the projections with respect to each other by applying the GRAPPA operator. The method was validated in numerical simulations, as well as in radial acquisitions of a phantom and in vivo images at 3T. The results of the correction were compared to a previously proposed correction method.

RESULTS

The newly introduced technique allowed for a reliable trajectory correction in each of the presented examples. The method was able to remove artifacts as effectively as methods that are based on data from additional calibration scans.

CONCLUSION

The iterative technique introduced in this paper allows for a correction of trajectory errors in radial imaging with no need for additional calibration data.

Algebraic reconstruction technique for parallel imaging reconstruction of undersampled radial data: application to cardiac cine

PURPOSE

To investigate algebraic reconstruction technique (ART) for parallel imaging reconstruction of radial data, applied to accelerated cardiac cine.

METHODS

A graphics processing unit (GPU)-accelerated ART reconstruction was implemented and applied to simulations, point spread functions and in 12 subjects imaged with radial cardiac cine acquisitions. Cine images were reconstructed with radial ART at multiple undersampling levels (192 Nr × Np  = 96 to 16). Images were qualitatively and quantitatively analyzed for sharpness and artifacts, and compared to filtered back-projection, and conjugate gradient SENSE.

RESULTS

Radial ART provided reduced artifacts and mainly preserved spatial resolution, for both simulations and in vivo data. Artifacts were qualitatively and quantitatively less with ART than filtered back-projection using 48, 32, and 24 Np , although filtered back-projection provided quantitatively sharper images at undersampling levels of 48-24 Np (all P < 0.05). Use of undersampled radial data for generating auto-calibrated coil-sensitivity profiles resulted in slightly reduced quality. ART was comparable to conjugate gradient SENSE. GPU-acceleration increased ART reconstruction speed 15-fold, with little impact on the images.

CONCLUSION

GPU-accelerated ART is an alternative approach to image reconstruction for parallel radial MR imaging, providing reduced artifacts while mainly maintaining sharpness compared to filtered back-projection, as shown by its first application in cardiac studies.