Highly efficient free-breathing 3D whole-heart imaging in 3-min: single center study in adults with congenital heart disease.

BACKGROUND: Three dimensional, whole-heart (3DWH) MRI is an established non-invasive imaging modality in patients with congenital heart disease (CHD) for the diagnosis of cardiovascular morphology and for clinical decision making. Current techniques utilise diaphragmatic navigation (dNAV) for respiratory motion correction and gating and are frequently limited by long acquisition times. This study proposes and evaluates the diagnostic performance of a respiratory gating-free framework, which considers respiratory image-based navigation (iNAV), and highly accelerated variable density Cartesian sampling in concert with non-rigid motion correction and low-rank patch-based denoising (iNAV-3DWH-PROST). The method is compared to the clinical dNAV-3DWH sequence in adult patients with CHD. METHODS: In this prospective single center study, adult patients with CHD who underwent the clinical dNAV-3DWH MRI were also scanned with the iNAV-3DWH-PROST. Diagnostic confidence (4-point Likert scale) and diagnostic accuracy for common cardiovascular lesions was assessed by three readers. Scan times and diagnostic confidence were compared using the Wilcoxon-signed rank test. Co-axial vascular dimensions at three anatomic landmarks were measured, and agreement between the research and the corresponding clinical sequence was assessed with Bland-Altman analysis. RESULTS: The study included 60 participants (mean age ± [SD]: 33 ± 14 years; 36 men). The mean acquisition time of iNAV-3DWH-PROST was significantly lower compared with the conventional clinical sequence (3.1 ± 0.9 min vs 13.9 ± 3.9 min, p < 0.0001). Diagnostic confidence was higher for the iNAV-3DWH-PROST sequence compared with the clinical sequence (3.9 ± 0.2 vs 3.4 ± 0.8, p < 0.001), however there was no significant difference in diagnostic accuracy. Narrow limits of agreement and mean bias less than 0.08 cm were found between the research and the clinical vascular measurements. CONCLUSIONS: The iNAV-3DWH-PROST framework provides efficient, high quality and robust 3D whole-heart imaging in significantly shorter scan time compared to the standard clinical sequence.

A motion-corrected deep-learning reconstruction framework for accelerating whole-heart magnetic resonance imaging in patients with congenital heart disease.

BACKGROUND: Cardiovascular magnetic resonance (CMR) is an important imaging modality for the assessment and management of adult patients with congenital heart disease (CHD). However, conventional techniques for three-dimensional (3D) whole-heart acquisition involve long and unpredictable scan times and methods that accelerate scans via k-space undersampling often rely on long iterative reconstructions. Deep-learning-based reconstruction methods have recently attracted much interest due to their capacity to provide fast reconstructions while often outperforming existing state-of-the-art methods. In this study, we sought to adapt and validate a non-rigid motion-corrected model-based deep learning (MoCo-MoDL) reconstruction framework for 3D whole-heart MRI in a CHD patient cohort. METHODS: The previously proposed deep-learning reconstruction framework MoCo-MoDL, which incorporates a non-rigid motion-estimation network and a denoising regularization network within an unrolled iterative reconstruction, was trained in an end-to-end manner using 39 CHD patient datasets. Once trained, the framework was evaluated in eight CHD patient datasets acquired with seven-fold prospective undersampling. Reconstruction quality was compared with the state-of-the-art non-rigid motion-corrected patch-based low-rank reconstruction method (NR-PROST) and against reference images (acquired with three-or-four-fold undersampling and reconstructed with NR-PROST). RESULTS: Seven-fold undersampled scan times were 2.1 ± 0.3 minutes and reconstruction times were ∼30 seconds, approximately 240 times faster than an NR-PROST reconstruction. Image quality comparable to the reference images was achieved using the proposed MoCo-MoDL framework, with no statistically significant differences found in any of the assessed quantitative or qualitative image quality measures. Additionally, expert image quality scores indicated the MoCo-MoDL reconstructions were consistently of a higher quality than the NR-PROST reconstructions of the same data, with the differences in 12 of the 22 scores measured for individual vascular structures found to be statistically significant. CONCLUSION: The MoCo-MoDL framework was applied to an adult CHD patient cohort, achieving good quality 3D whole-heart images from ∼2-minute scans with reconstruction times of ∼30 seconds.

Magnetization Transfer BOOST Noncontrast Angiography Improves Pulmonary Vein Imaging in Adults With Congenital Heart Disease.

BACKGROUND: Cardiac MRI plays an important role in the diagnosis and follow-up of patients with congenital heart disease (CHD). Gadolinium-based contrast agents are often needed to overcome flow-related and off-resonance artifacts that can impair the quality of conventional noncontrast 3D imaging. As serial imaging is often required in CHD, the development of robust noncontrast 3D MRI techniques is desirable. PURPOSE: To assess the clinical utility of noncontrast enhanced magnetization transfer and inversion recovery prepared 3D free-breathing sequence (MTC-BOOST) compared to conventional 3D whole heart imaging in patients with CHD. STUDY TYPE: Prospective, image quality. POPULATION: A total of 27 adult patients (44% female, mean age 30.9 ± 14.8 years) with CHD. FIELD STRENGTH/SEQUENCE: A 1.5 T; free-breathing 3D MTC-BOOST sequence. ASSESSMENT: MTC-BOOST was compared to diaphragmatic navigator-gated, noncontrast T2 prepared 3D whole-heart imaging sequence (T2prep-3DWH) for comparison of vessel dimensions, lumen-to-myocardium contrast ratio (CR), and image quality (vessel wall sharpness and presence and type of artifacts) assessed by two experienced cardiologists on a 5-point scale. STATISTICAL TESTS: Mann-Whitney test, paired Wilcoxon signed-rank test, Bland-Altman plots. P < 0.05 was considered statistically significant. RESULTS: MTC-BOOST significantly improved image quality and CR of the right-sided pulmonary veins (PV): (CR: right upper PV 1.06 ± 0.50 vs. 0.58 ± 0.74; right lower PV 1.32 ± 0.38 vs. 0.81 ± 0.73) compared to conventional T2prep-3DWH imaging where the PVs were not visualized in some cases due to off-resonance effects. MTC-BOOST demonstrated resistance to degradation of luminal signal (assessed by CR) secondary to accelerated or turbulent flow conditions. T2prep-3DWH had higher image quality scores than MTC-BOOST for the aorta and coronary arteries; however, great vessel dimensions derived from MTC-BOOST showed excellent agreement with standard T2prep-3DWH imaging. DATA CONCLUSION: MTC-BOOST allows for improved contrast-free imaging of pulmonary veins and regions characterized by accelerated or turbulent blood flow compared to standard T2prep-3DWH imaging, with excellent agreement of great vessel dimensions. EVIDENCE LEVEL: 1 TECHNICAL EFFICACY: Stage 2.

Deep-learning based super-resolution for 3D isotropic coronary MR angiography in less than a minute.

PURPOSE: To develop and evaluate a novel and generalizable super-resolution (SR) deep-learning framework for motion-compensated isotropic 3D coronary MR angiography (CMRA), which allows free-breathing acquisitions in less than a minute. METHODS: Undersampled motion-corrected reconstructions have enabled free-breathing isotropic 3D CMRA in ~5-10 min acquisition times. In this work, we propose a deep-learning-based SR framework, combined with non-rigid respiratory motion compensation, to shorten the acquisition time to less than 1 min. A generative adversarial network (GAN) is proposed consisting of two cascaded Enhanced Deep Residual Network generator, a trainable discriminator, and a perceptual loss network. A 16-fold increase in spatial resolution is achieved by reconstructing a high-resolution (HR) isotropic CMRA (0.9 mm3 or 1.2 mm3 ) from a low-resolution (LR) anisotropic CMRA (0.9 × 3.6 × 3.6 mm3 or 1.2 × 4.8 × 4.8 mm3 ). The impact and generalization of the proposed SRGAN approach to different input resolutions and operation on image and patch-level is investigated. SRGAN was evaluated on a retrospective downsampled cohort of 50 patients and on 16 prospective patients that were scanned with LR-CMRA in ~50 s under free-breathing. Vessel sharpness and length of the coronary arteries from the SR-CMRA is compared against the HR-CMRA. RESULTS: SR-CMRA showed statistically significant (P < .001) improved vessel sharpness 34.1% ± 12.3% and length 41.5% ± 8.1% compared with LR-CMRA. Good generalization to input resolution and image/patch-level processing was found. SR-CMRA enabled recovery of coronary stenosis similar to HR-CMRA with comparable qualitative performance. CONCLUSION: The proposed SR-CMRA provides a 16-fold increase in spatial resolution with comparable image quality to HR-CMRA while reducing the predictable scan time to <1 min.

3D whole-heart grey-blood late gadolinium enhancement cardiovascular magnetic resonance imaging.

PURPOSE: To develop a free-breathing whole-heart isotropic-resolution 3D late gadolinium enhancement (LGE) sequence with Dixon-encoding, which provides co-registered 3D grey-blood phase-sensitive inversion-recovery (PSIR) and complementary 3D fat volumes in a single scan of < 7 min. METHODS: A free-breathing 3D PSIR LGE sequence with dual-echo Dixon readout with a variable density Cartesian trajectory with acceleration factor of 3 is proposed. Image navigators are acquired to correct both inversion recovery (IR)-prepared and reference volumes for 2D translational respiratory motion, enabling motion compensated PSIR reconstruction with 100% respiratory scan efficiency. An intermediate PSIR reconstruction is performed between the in-phase echoes to estimate the signal polarity which is subsequently applied to the IR-prepared water volume to generate a water grey-blood PSIR image. The IR-prepared water volume is obtained using a water/fat separation algorithm from the corresponding dual-echo readout. The complementary fat-volume is obtained after water/fat separation of the reference volume. Ten patients (6 with myocardial scar) were scanned with the proposed water/fat grey-blood 3D PSIR LGE sequence at 1.5 T and compared to breath-held grey-blood 2D LGE sequence in terms of contrast ratio (CR), contrast-to-noise ratio (CNR), scar depiction, scar transmurality, scar mass and image quality. RESULTS: Comparable CRs (p = 0.98, 0.40 and 0.83) and CNRs (p = 0.29, 0.40 and 0.26) for blood-myocardium, scar-myocardium and scar-blood respectively were obtained with the proposed free-breathing 3D water/fat LGE and 2D clinical LGE scan. Excellent agreement for scar detection, scar transmurality, scar mass (bias = 0.29%) and image quality scores (from 1: non-diagnostic to 4: excellent) of 3.8 ± 0.42 and 3.6 ± 0.69 (p > 0.99) were obtained with the 2D and 3D PSIR LGE approaches with comparable total acquisition time (p = 0.29). Similar agreement in intra and inter-observer variability were obtained for the 2D and 3D acquisition respectively. CONCLUSION: The proposed approach enabled the acquisition of free-breathing motion-compensated isotropic-resolution 3D grey-blood PSIR LGE and fat volumes. The proposed approach showed good agreement with conventional 2D LGE in terms of CR, scar depiction and scan time, while enabling free-breathing acquisition, whole-heart coverage, reformatting in arbitrary views and visualization of both water and fat information.

3D whole-heart isotropic-resolution motion-compensated joint T1 /T2 mapping and water/fat imaging.

PURPOSE: To develop a free-breathing isotropic-resolution whole-heart joint T1 and T2 mapping sequence with Dixon-encoding that provides coregistered 3D T1 and T2 maps and complementary 3D anatomical water and fat images in a single ~9 min scan. METHODS: Four interleaved dual-echo Dixon gradient echo volumes are acquired with a variable density Cartesian trajectory and different preparation pulses: 1) inversion recovery-preparation, 2) and 3) no preparations, and 4) T2 preparation. Image navigators are acquired to correct each echo for 2D translational respiratory motion; the 8 echoes are jointly reconstructed with a low-rank patch-based reconstruction. A water/fat separation algorithm is used to obtain water and fat images for each acquired volume. T1 and T2 maps are generated by matching the signal evolution of the water images to a simulated dictionary. Complementary bright-blood and fat volumes for anatomical visualization are obtained from the T2 -prepared dataset. The proposed sequence was tested in phantom experiments and 10 healthy subjects and compared to standard 2D MOLLI T1 mapping, 2D balance steady-state free precession T2 mapping, and 3D T2 -prepared Dixon coronary MR angiography. RESULTS: High linear correlation was found between T1 and T2 quantification with the proposed approach and phantom spin echo measurements (y = 1.1 × -11.68, R2 = 0.98; and y = 0.85 × +5.7, R2 = 0.99). Mean myocardial values of T1 /T2 = 1116 ± 30.5 ms/45.1 ± 2.38 ms were measured in vivo. Biases of T1 /T2 = 101.8 ms/-0.77 ms were obtained compared to standard 2D techniques. CONCLUSION: The proposed joint T1 /T2 sequence permitted the acquisition of motion-compensated isotropic-resolution 3D T1 and T2 maps and complementary coronary MR angiography and fat volumes, showing promising results in terms of T1 and T2 quantification and visualization of cardiac anatomy and pericardial fat.

3D Whole-heart free-breathing qBOOST-T2 mapping.

PURPOSE: To develop an accelerated motion corrected 3D whole-heart imaging approach (qBOOST-T2) for simultaneous high-resolution bright- and black-blood cardiac MR imaging and quantitative myocardial T2 characterization. METHODS: Three undersampled interleaved balanced steady-state free precession cardiac MR volumes were acquired with a variable density Cartesian trajectory and different magnetization preparations: (1) T2-prepared inversion recovery (T2prep-IR), (2) T2-preparation, and (3) no preparation. Image navigators were acquired prior the acquisition to correct for 2D translational respiratory motion. Each 3D volume was reconstructed with a low-rank patch-based reconstruction. The T2prep-IR volume provides bright-blood anatomy visualization, the black-blood volume is obtained by means of phase sensitive reconstruction between first and third datasets, and T2 maps are generated by matching the signal evolution to a simulated dictionary. The proposed sequence has been evaluated in simulations, phantom experiments, 11 healthy subjects and compared with 3D bright-blood cardiac MR and standard 2D breath-hold balanced steady-state free precession T2 mapping. The feasibility of the proposed approach was tested on 4 patients with suspected cardiovascular disease. RESULTS: High linear correlation (y = 1.09 × -0.83, R2 = 0.99) was found between the proposed qBOOST-T2 and T2 spin echo measurements in phantom experiment. Good image quality was observed in vivo with the proposed 4x undersampled qBOOST-T2. Mean T2 values of 53.1 ± 2.1 ms and 55.8 ± 2.7 ms were measured in vivo for 2D balanced steady-state free precession T2 mapping and qBOOST-T2, respectively, with linear correlation of y = 1.02x+1.46 (R2 = 0.61) and T2 bias = 2.7 ms. CONCLUSION: The proposed qBOOST-T2 sequence allows the acquisition of 3D high-resolution co-registered bright- and black-blood volumes and T2 maps in a single scan of ~11 min, showing promising results in terms of T2 quantification.

3D free-breathing cardiac magnetic resonance fingerprinting.

PURPOSE: To develop a novel respiratory motion compensated three-dimensional (3D) cardiac magnetic resonance fingerprinting (cMRF) approach for whole-heart myocardial T1 and T2 mapping from a free-breathing scan. METHODS: Two-dimensional (2D) cMRF has been recently proposed for simultaneous, co-registered T1 and T2 mapping from a breath-hold scan; however, coverage is limited. Here we propose a novel respiratory motion compensated 3D cMRF approach for whole-heart myocardial T1 and T2 tissue characterization from a free-breathing scan. Variable inversion recovery and T2 preparation modules are used for parametric encoding, respiratory bellows driven localized autofocus is proposed for beat-to-beat translation motion correction and a subspace regularized reconstruction is employed to accelerate the scan. The proposed 3D cMRF approach was evaluated in a standardized T1 /T2 phantom in comparison with reference spin echo values and in 10 healthy subjects in comparison with standard 2D MOLLI, SASHA and T2 -GraSE mapping techniques at 1.5 T. RESULTS: 3D cMRF T1 and T2 measurements were generally in good agreement with reference spin echo values in the phantom experiments, with relative errors of 2.9% and 3.8% for T1 and T2 (T2 < 100 ms), respectively. in vivo left ventricle (LV) myocardial T1 values were 1054 ± 19 ms for MOLLI, 1146 ± 20 ms for SASHA and 1093 ± 24 ms for the proposed 3D cMRF; corresponding T2 values were 51.8 ± 1.6 ms for T2-GraSE and 44.6 ± 2.0 ms for 3D cMRF. LV coefficients of variation were 7.6 ± 1.6% for MOLLI, 12.1 ± 2.7% for SASHA and 5.8 ± 0.8% for 3D cMRF T1 , and 10.5 ± 1.4% for T2-GraSE and 11.7 ± 1.6% for 3D cMRF T2 . CONCLUSION: The proposed 3D cMRF can provide whole-heart, simultaneous and co-registered T1 and T2 maps with accuracy and precision comparable to those of clinical standards in a single free-breathing scan of about 7 min.

Motion-corrected 3D whole-heart water-fat high-resolution late gadolinium enhancement cardiovascular magnetic resonance imaging.

BACKGROUND: Conventional 2D inversion recovery (IR) and phase sensitive inversion recovery (PSIR) late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) have been widely incorporated into routine CMR for the assessment of myocardial viability. However, reliable suppression of fat signal, and increased isotropic spatial resolution and volumetric coverage within a clinically feasible scan time remain a challenge. In order to address these challenges, this work proposes a highly efficient respiratory motion-corrected 3D whole-heart water/fat LGE imaging framework. METHODS: An accelerated IR-prepared 3D dual-echo acquisition and motion-corrected reconstruction framework for whole-heart water/fat LGE imaging was developed. The acquisition sequence includes 2D image navigators (iNAV), which are used to track the respiratory motion of the heart and enable 100% scan efficiency. Non-rigid motion information estimated from the 2D iNAVs and from the data itself is integrated into a high-dimensional patch-based undersampled reconstruction technique (HD-PROST), to produce high-resolution water/fat 3D LGE images. A cohort of 20 patients with known or suspected cardiovascular disease was scanned with the proposed 3D water/fat LGE approach. 3D water LGE images were compared to conventional breath-held 2D LGE images (2-chamber, 4-chamber and stack of short-axis views) in terms of image quality (1: full diagnostic to 4: non-diagnostic) and presence of LGE findings. RESULTS: Image quality was considered diagnostic in 18/20 datasets for both 2D and 3D LGE magnitude images, with comparable image quality scores (2D: 2.05 ± 0.72, 3D: 1.88 ± 0.90, p-value = 0.62) and overall agreement in LGE findings. Acquisition time for isotropic high-resolution (1.3mm3) water/fat LGE images was 8.0 ± 1.4 min (3-fold acceleration, 60-88 slices covering the whole heart), while 2D LGE images were acquired in 5.6 ± 2.2 min (12-18 slices, including pauses between breath-holds) albeit with a lower spatial resolution (1.40-1.75 mm in-plane × 8 mm slice thickness). CONCLUSION: A novel framework for motion-corrected whole-heart 3D water/fat LGE imaging has been introduced. The method was validated in patients with known or suspected cardiovascular disease, showing good agreement with conventional breath-held 2D LGE imaging, but offering higher spatial resolution, improved volumetric coverage and good image quality from a free-breathing acquisition with 100% scan efficiency and predictable scan time.

Simultaneous 3D whole-heart bright-blood and black blood imaging for cardiovascular anatomy and wall assessment with interleaved T2 prep-IR.

PURPOSE: To develop a motion-corrected 3D flow-insensitive imaging approach interleaved T2 prepared-inversion recovery (iT2 prep-IR) for simultaneous lumen and wall visualization of the great thoracic vessels and cardiac structures. METHODS: A 3D flow-insensitive approach for simultaneous cardiovascular lumen and wall visualization (iT2 prep) has been previously proposed. This approach requires subject-dependent weighted subtraction to completely null the arterial blood signal in the black-blood volume. Here, we propose an (T2 prep-IR) approach to improve wall visualization and remove need for weighted subtraction. The proposed sequence is based on the acquisition and direct subtraction of 2 interleaved 3D whole-heart data sets acquired with and without T2 prep-IR preparation. Image navigators are acquired before data acquisition to enable 2D translational and 3D non-rigid motion correction allowing 100% respiratory scan efficiency. The proposed approach was evaluated in 10 healthy subjects and compared with the conventional 2D double inversion recovery (DIR) sequence and the 3D iT2 prep sequence. Additionally, 5 patients with congenital heart disease were acquired to test the clinical feasibility of the proposed approach. RESULTS: The proposed iT2 prep-IR sequence showed improved blood nulling compared to both DIR and iT2 prep techniques in terms of SNR (SNRblood = 6.9, 12.2, and 18.2, respectively) and contrast-to-noise-ratio (CNRmyoc-blood = 28.4, 15.4, and 15.3, respectively). No statistical difference was observed between iT2 prep-IR, iT2 prep and DIR atrial and ventricular wall thickness quantification. CONCLUSION: The proposed interleaved T2 prep-IR sequence enables the simultaneous lumen and wall visualization of cardiac structures and shows promising results in terms of SNR, CNR, and wall thickness measurement.

The feasibility of relaxation-enhanced angiography without contrast and triggering for preprocedural planning of transcatheter aortic valve implantation.

Background: Cardiovascular MRI is advantageous in transcatheter aortic valve implantation (TAVI) planning. This study aimed to evaluate the feasibility of comprehensive non-contrast MRI [relaxation-enhanced angiography without contrast and triggering (REACT)] combined with a three-dimensional whole-heart MRI protocol for preprocedural planning of TAVI vs. computed tomography angiography (CTA). Methods: Thirty patients with severe aortic stenosis were prospectively enrolled. The anatomical properties of the aortic root anatomy, including the perimeter and area of the virtual aortic valve annulus and coronary heights, were determined from 3D whole-heart MRI and cardiac CTA (CCTA) images, respectively. The diameters of the aorta (thoracic and abdominal aorta) and iliofemoral arteries were measured from REACT and aortic CTA (ACTA) images, respectively. A paired t-test was used to compare these two modalities. Bland-Altman plots were used to assess cardiovascular MRI and CTA measurements. Transcatheter heart valve (THV) sizing was performed based on CCTA measurements and compared with 3D whole-heart MRI measurements. The extent of annular calcification on 3D whole-heart MRI images was evaluated by a four-point grading scale and compared with CCTA data. Results: All 30 patients completed CTA and cardiovascular MRI examinations, with the TAVI procedure being administered in 25 patients. The mean acquisition time of the comprehensive MRI protocol was 18 ± 3.2 min. There were no significant differences between ACTA and REACT data in regard to the diameters of aortic and iliofemoral arteries, including the ascending thoracic aorta (37 ± 4.6 mm vs. 37.7 ± 5.2 mm, p = 0.085), descending thoracic aorta (24.3 ± 2.8 mm vs. 24.3 ± 2.8 mm, p = 0.832), abdominal aorta (20.9 ± 2.5 mm vs. 20.8 ± 2.5 mm, p = 0.602), bilateral common iliac arteries (right: 8.36 ± 1.44 mm vs. 8.42 ± 1.27 mm, p = 0.590; left: 8.61 ± 1.71 mm vs. 8.86 ± 1.46 mm, p = 0.050), and bilateral femoral arteries (right: 6.77 ± 1.06 mm vs. 6.87 ± 1.00 mm, p = 0.157; left: 6.75 ± 1.02 mm vs. 6.90 ± 0.80 mm, p = 0.142). Both modalities showed similar aortic valve morphology and semi-quantitative valve calcification (all, p's > 0.05). Overall agreement for implanted THV was found in all 25 (100%) patients assessed with both modalities. Conclusion: REACT combined with 3D whole-heart MRI enables reliable measurements of aortic root anatomy, annular calcification, and aorta and iliofemoral access in patients under evaluation for TAVI.

Accelerating 3D MTC-BOOST in patients with congenital heart disease using a joint multi-scale variational neural network reconstruction.

PURPOSE: Free-breathing Magnetization Transfer Contrast Bright blOOd phase SensiTive (MTC-BOOST) is a prototype balanced-Steady-State Free Precession sequence for 3D whole-heart imaging, that employs the endogenous magnetisation transfer contrast mechanism. This achieves reduction of flow and off-resonance artefacts, that often arise with the clinical T2prepared balanced-Steady-State Free Precession sequence, enabling high quality, contrast-agent free imaging of the thoracic cardiovascular anatomy. Fully-sampled MTC-BOOST acquisition requires long scan times (~10-24 min) and therefore acceleration is needed to permit its clinical incorporation. The aim of this study is to enable and clinically validate the 5-fold accelerated MTC-BOOST acquisition with joint Multi-Scale Variational Neural Network (jMS-VNN) reconstruction. METHODS: Thirty-six patients underwent free-breathing, 3D whole-heart imaging with the MTC-BOOST sequence, which is combined with variable density spiral-like Cartesian sampling and 2D image navigators for translational motion estimation. This sequence acquires two differently weighted bright-blood volumes in an interleaved fashion, which are then joined in a phase sensitive inversion recovery reconstruction to obtain a complementary fully co-registered black-blood volume. Data from eighteen patients were used for training, whereas data from the remaining eighteen patients were used for testing/evaluation. The proposed deep-learning based approach adopts a supervised multi-scale variational neural network for joint reconstruction of the two differently weighted bright-blood volumes acquired with the 5-fold accelerated MTC-BOOST. The two contrast images are stacked as different channels in the network to exploit the shared information. The proposed approach is compared to the fully-sampled MTC-BOOST and 5-fold undersampled MTC-BOOST acquisition with Compressed Sensing (CS) reconstruction in terms of scan/reconstruction time and bright-blood image quality. Comparison against conventional 2-fold undersampled T2-prepared 3D bright-blood whole-heart clinical sequence (T2prep-3DWH) is also included. RESULTS: Acquisition time was 3.0 ±â€¯1.0 min for the 5-fold accelerated MTC-BOOST versus 9.0 ±â€¯1.1 min for the fully-sampled MTC-BOOST and 11.1 ±â€¯2.6 min for the T2prep-3DWH (p < 0.001 and p < 0.001, respectively). Reconstruction time was significantly lower with the jMS-VNN method compared to CS (10 ±â€¯0.5 min vs 20 ±â€¯2 s, p < 0.001). Image quality was higher for the proposed 5-fold undersampled jMS-VNN versus conventional CS, comparable or higher to the corresponding T2prep-3DWH dataset and similar to the fully-sampled MTC-BOOST. CONCLUSION: The proposed 5-fold accelerated jMS-VNN MTC-BOOST framework provides efficient 3D whole-heart bright-blood imaging in fast acquisition and reconstruction time with concomitant reduction of flow and off-resonance artefacts, that are frequently encountered with the clinical sequence. Image quality of the cardiac anatomy and thoracic vasculature is comparable or superior to the clinical scan and 5-fold CS reconstruction in faster reconstruction time, promising potential clinical adoption.

Self-supervised learning-based diffeomorphic non-rigid motion estimation for fast motion-compensated coronary MR angiography.

PURPOSE: To accelerate non-rigid motion corrected coronary MR angiography (CMRA) reconstruction by developing a deep learning based non-rigid motion estimation network and combining this with an efficient implementation of the undersampled motion corrected reconstruction. METHODS: Undersampled and respiratory motion corrected CMRA with overall short scans of 5 to 10 min have been recently proposed. However, image reconstruction with this approach remains lengthy, since it relies on several non-rigid image registrations to estimate the respiratory motion and on a subsequent iterative optimization to correct for motion during the undersampled reconstruction. Here we introduce a self-supervised diffeomorphic non-rigid respiratory motion estimation network, DiRespME-net, to speed up respiratory motion estimation. We couple this with an efficient GPU-based implementation of the subsequent motion-corrected iterative reconstruction. DiRespME-net is based on a U-Net architecture, and is trained in a self-supervised fashion, with a loss enforcing image similarity and spatial smoothness of the motion fields. Motion predicted by DiRespME-net was used for GPU-based motion-corrected CMRA in 12 test subjects and final images were compared to those produced by state-of-the-art reconstruction. Vessel sharpness and visible length of the right coronary artery (RCA) and the left anterior descending (LAD) coronary artery were used as metrics of image quality for comparison. RESULTS: No statistically significant difference in image quality was found between images reconstructed with the proposed approach (MC:DiRespME-net) and a motion-corrected reconstruction using cubic B-splines (MC:Nifty-reg). Visible vessel length was not significantly different between methods (RCA: MC:Nifty-reg 5.7 ± 1.7 cm vs MC:DiRespME-net 5.8 ± 1.7 cm, P = 0.32; LAD: MC:Nifty-reg 7.0 ± 2.6 cm vs MC:DiRespME-net 6.9 ± 2.7 cm, P = 0.81). Similarly, no statistically significant difference between methods was observed in terms of vessel sharpness (RCA: MC:Nifty-reg 60.3 ± 7.2% vs MC:DiRespME-net 61.0 ± 6.8%, P = 0.19; LAD: MC:Nifty-reg 57.4 ± 7.9% vs MC:DiRespME-net 58.1 ± 7.5%, P = 0.27). The proposed approach achieved a 50-fold reduction in computation time, resulting in a total reconstruction time of approximately 20 s. CONCLUSIONS: The proposed self-supervised learning-based motion corrected reconstruction enables fast motion-corrected CMRA image reconstruction, holding promise for integration in clinical routine.

Self-navigated 3D whole-heart MRA for non-enhanced surveillance of thoracic aortic dilation: A comparison to CTA.

PURPOSE: To prospectively compare image quality and reliability of a non-contrast, self-navigated 3D whole-heart magnetic resonance angiography (MRA) sequence with contrast-enhanced computed tomography angiography (CTA) for sizing of thoracic aortic aneurysm (TAA). METHODS: Self-navigated 3D whole-heart 1.5 T MRA was performed in 20 patients (aged 67 ± 9 years, 75% male) for sizing of TAA; a subgroup of 18 (90%) patients underwent additional contrast-enhanced CTA on the same day. Subjective image quality was scored according to a 4-point Likert scale and ratings between observers were compared by Cohen's Kappa statistics. For MRA, subjective motion blurring and signal inhomogeneity was rated according to a 3-point scale, respectively. Objective signal inhomogeneity of MRA was quantified as standard deviation of the voxel intensities in a circular region of interest (ROI) placed in the ascending aorta divided by their mean value. Continuous MRA and CTA measurements were analyzed with regression and Bland-Altman analysis. RESULTS: Overall subjective image quality as rated by two observers was 1 [interquartile range (IQR) 1-2] for self-navigated MRA and 1.5 [IQR 1-2] for CTA (p = 0.717). For MRA, perfect inter-observer agreement was found regarding presence of artefacts and subjective image sharpness (κ = 1). Subjective signal inhomogeneity agreed moderately between the observers (κ = 0.58, p = 0.007), however, it correlated strongly with objectively quantified inhomogeneity of the blood pool signal (r = 0.78, p < 0.0001). Maximum diameters of TAA as measured by self-navigated MRA and CTA showed very strong correlation (r = 0.99, p < 0.0001) without significant inter-method bias (bias -0.03 mm, lower and upper limit of agreement -0.74 and 0.68 mm, p = 0.749). Inter-observer correlation of aortic aneurysm as measured by MRA was very strong (r = 0.96) without significant bias (p = 0.695). CONCLUSION: Self-navigated 3D whole-heart MRA enables reliable contrast- and radiation free aortic dilation surveillance without significant difference to standardized CTA while providing predictable acquisition time and offering excellent image quality.