Training deep learning based dynamic MR image reconstruction using open-source natural videos.

To develop and assess a deep learning (DL) pipeline to learn dynamic MR image reconstruction from publicly available natural videos (Inter4K). Learning was performed for a range of DL architectures (VarNet, 3D UNet, FastDVDNet) and corresponding sampling patterns (Cartesian, radial, spiral) either from true multi-coil cardiac MR data (N = 692) or from synthetic MR data simulated from Inter4K natural videos (N = 588). Real-time undersampled dynamic MR images were reconstructed using DL networks trained with cardiac data and natural videos, and compressed sensing (CS). Differences were assessed in simulations (N = 104 datasets) in terms of MSE, PSNR, and SSIM and prospectively for cardiac cine (short axis, four chambers, N = 20) and speech cine (N = 10) data in terms of subjective image quality ranking, SNR and Edge sharpness. Friedman Chi Square tests with post-hoc Nemenyi analysis were performed to assess statistical significance. In simulated data, DL networks trained with cardiac data outperformed DL networks trained with natural videos, both of which outperformed CS (p < 0.05). However, in prospective experiments DL reconstructions using both training datasets were ranked similarly (and higher than CS) and presented no statistical differences in SNR and Edge Sharpness for most conditions.The developed pipeline enabled learning dynamic MR reconstruction from natural videos preserving DL reconstruction advantages such as high quality fast and ultra-fast reconstructions while overcoming some limitations (data scarcity or sharing). The natural video dataset, code and pre-trained networks are made readily available on github.

HyperSLICE: HyperBand optimized spiral for low-latency interactive cardiac examination.

PURPOSE: Interactive cardiac MRI is used for fast scan planning and MR-guided interventions. However, the requirement for real-time acquisition and near-real-time visualization constrains the achievable spatio-temporal resolution. This study aims to improve interactive imaging resolution through optimization of undersampled spiral sampling and leveraging of deep learning for low-latency reconstruction (deep artifact suppression). METHODS: A variable density spiral trajectory was parametrized and optimized via HyperBand to provide the best candidate trajectory for rapid deep artifact suppression. Training data consisted of 692 breath-held CINEs. The developed interactive sequence was tested in simulations and prospectively in 13 subjects (10 for image evaluation, 2 during catheterization, 1 during exercise). In the prospective study, the optimized framework-HyperSLICE- was compared with conventional Cartesian real-time and breath-hold CINE imaging in terms quantitative and qualitative image metrics. Statistical differences were tested using Friedman chi-squared tests with post hoc Nemenyi test (p < 0.05). RESULTS: In simulations the normalized RMS error, peak SNR, structural similarity, and Laplacian energy were all statistically significantly higher using optimized spiral compared to radial and uniform spiral sampling, particularly after scan plan changes (structural similarity: 0.71 vs. 0.45 and 0.43). Prospectively, HyperSLICE enabled a higher spatial and temporal resolution than conventional Cartesian real-time imaging. The pipeline was demonstrated in patients during catheter pull back, showing sufficiently fast reconstruction for interactive imaging. CONCLUSION: HyperSLICE enables high spatial and temporal resolution interactive imaging. Optimizing the spiral sampling enabled better overall image quality and superior handling of image transitions compared with radial and uniform spiral trajectories.

2D sodium MRI of the human calf using half-sinc excitation pulses and compressed sensing.

PURPOSE: Sodium MRI can be used to quantify tissue sodium concentration (TSC) in vivo; however, UTE sequences are required to capture the rapidly decaying signal. 2D MRI enables high in-plane resolution but typically has long TEs. Half-sinc excitation may enable UTE; however, twice as many readouts are necessary. Scan time can be minimized by reducing the number of signal averages (NSAs), but at a cost to SNR. We propose using compressed sensing (CS) to accelerate 2D half-sinc acquisitions while maintaining SNR and TSC. METHODS: Ex vivo and in vivo TSC were compared between 2D spiral sequences with full-sinc (TE = 0.73 ms, scan time ≈ 5 min) and half-sinc excitation (TE = 0.23 ms, scan time ≈ 10 min), with 150 NSAs. Ex vivo, these were compared to a reference 3D sequence (TE = 0.22 ms, scan time ≈ 24 min). To investigate shortening 2D scan times, half-sinc data was retrospectively reconstructed with fewer NSAs, comparing a nonuniform fast Fourier transform to CS. Resultant TSC and image quality were compared to reference 150 NSAs nonuniform fast Fourier transform images. RESULTS: TSC was significantly higher from half-sinc than from full-sinc acquisitions, ex vivo and in vivo. Ex vivo, half-sinc data more closely matched the reference 3D sequence, indicating improved accuracy. In silico modeling confirmed this was due to shorter TEs minimizing bias caused by relaxation differences between phantoms and tissue. CS was successfully applied to in vivo, half-sinc data, maintaining TSC and image quality (estimated SNR, edge sharpness, and qualitative metrics) with ≥50 NSAs. CONCLUSION: 2D sodium MRI with half-sinc excitation and CS was validated, enabling TSC quantification with 2.25 × 2.25 mm2 resolution and scan times of ≤5 mins.

Shape-driven deep neural networks for fast acquisition of aortic 3D pressure and velocity flow fields.

Computational fluid dynamics (CFD) can be used to simulate vascular haemodynamics and analyse potential treatment options. CFD has shown to be beneficial in improving patient outcomes. However, the implementation of CFD for routine clinical use is yet to be realised. Barriers for CFD include high computational resources, specialist experience needed for designing simulation set-ups, and long processing times. The aim of this study was to explore the use of machine learning (ML) to replicate conventional aortic CFD with automatic and fast regression models. Data used to train/test the model consisted of 3,000 CFD simulations performed on synthetically generated 3D aortic shapes. These subjects were generated from a statistical shape model (SSM) built on real patient-specific aortas (N = 67). Inference performed on 200 test shapes resulted in average errors of 6.01% ±3.12 SD and 3.99% ±0.93 SD for pressure and velocity, respectively. Our ML-based models performed CFD in ∼0.075 seconds (4,000x faster than the solver). This proof-of-concept study shows that results from conventional vascular CFD can be reproduced using ML at a much faster rate, in an automatic process, and with reasonable accuracy.

Automatic segmentation of the great arteries for computational hemodynamic assessment.

BACKGROUND: Computational fluid dynamics (CFD) is increasingly used for the assessment of blood flow conditions in patients with congenital heart disease (CHD). This requires patient-specific anatomy, typically obtained from segmented 3D cardiovascular magnetic resonance (CMR) images. However, segmentation is time-consuming and requires expert input. This study aims to develop and validate a machine learning (ML) method for segmentation of the aorta and pulmonary arteries for CFD studies. METHODS: 90 CHD patients were retrospectively selected for this study. 3D CMR images were manually segmented to obtain ground-truth (GT) background, aorta and pulmonary artery labels. These were used to train and optimize a U-Net model, using a 70-10-10 train-validation-test split. Segmentation performance was primarily evaluated using Dice score. CFD simulations were set up from GT and ML segmentations using a semi-automatic meshing and simulation pipeline. Mean pressure and velocity fields across 99 planes along the vessel centrelines were extracted, and a mean average percentage error (MAPE) was calculated for each vessel pair (ML vs GT). A second observer (SO) segmented the test dataset for assessment of inter-observer variability. Friedman tests were used to compare ML vs GT, SO vs GT and ML vs SO metrics, and pressure/velocity field errors. RESULTS: The network's Dice score (ML vs GT) was 0.945 (interquartile range: 0.929-0.955) for the aorta and 0.885 (0.851-0.899) for the pulmonary arteries. Differences with the inter-observer Dice score (SO vs GT) and ML vs SO Dice scores were not statistically significant for either aorta or pulmonary arteries (p = 0.741, p = 0.061). The ML vs GT MAPEs for pressure and velocity in the aorta were 10.1% (8.5-15.7%) and 4.1% (3.1-6.9%), respectively, and for the pulmonary arteries 14.6% (11.5-23.2%) and 6.3% (4.3-7.9%), respectively. Inter-observer (SO vs GT) and ML vs SO pressure and velocity MAPEs were of a similar magnitude to ML vs GT (p > 0.2). CONCLUSIONS: ML can successfully segment the great vessels for CFD, with errors similar to inter-observer variability. This fast, automatic method reduces the time and effort needed for CFD analysis, making it more attractive for routine clinical use.

FReSCO: Flow Reconstruction and Segmentation for low-latency Cardiac Output monitoring using deep artifact suppression and segmentation.

PURPOSE: Real-time monitoring of cardiac output (CO) requires low-latency reconstruction and segmentation of real-time phase-contrast MR, which has previously been difficult to perform. Here we propose a deep learning framework for "FReSCO" (Flow Reconstruction and Segmentation for low latency Cardiac Output monitoring). METHODS: Deep artifact suppression and segmentation U-Nets were independently trained. Breath-hold spiral phase-contrast MR data (N = 516) were synthetically undersampled using a variable-density spiral sampling pattern and gridded to create aliased data for training of the artifact suppression U-net. A subset of the data (N = 96) was segmented and used to train the segmentation U-net. Real-time spiral phase-contrast MR was prospectively acquired and then reconstructed and segmented using the trained models (FReSCO) at low latency at the scanner in 10 healthy subjects during rest, exercise, and recovery periods. Cardiac output obtained via FReSCO was compared with a reference rest CO and rest and exercise compressed-sensing CO. RESULTS: The FReSCO framework was demonstrated prospectively at the scanner. Beat-to-beat heartrate, stroke volume, and CO could be visualized with a mean latency of 622 ms. No significant differences were noted when compared with reference at rest (bias = -0.21 ± 0.50 L/min, p = 0.246) or compressed sensing at peak exercise (bias = 0.12 ± 0.48 L/min, p = 0.458). CONCLUSIONS: The FReSCO framework was successfully demonstrated for real-time monitoring of CO during exercise and could provide a convenient tool for assessment of the hemodynamic response to a range of stressors.

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

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

Real-time deep artifact suppression using recurrent U-Nets for low-latency cardiac MRI.

PURPOSE: Real-time low latency MRI is performed to guide various cardiac interventions. Real-time acquisitions often require iterative image reconstruction strategies, which lead to long reconstruction times. In this study, we aim to reconstruct highly undersampled radial real-time data with low latency using deep learning. METHODS: A 2D U-Net with convolutional long short-term memory layers is proposed to exploit spatial and preceding temporal information to reconstruct highly accelerated tiny golden radial data with low latency. The network was trained using a dataset of breath-hold CINE data (including 770 time series from 7 different orientations). Synthetic paired data were created by retrospectively undersampling the magnitude images, and the network was trained to recover the target images. In the spirit of interventional imaging, the network was trained and tested for varying acceleration rates and orientations. Data were prospectively acquired and reconstructed in real time in 1 healthy subject interactively and in 3 patients who underwent catheterization. Images were visually compared to sliding window and compressed sensing reconstructions and a conventional Cartesian real-time sequence. RESULTS: The proposed network generalized well to different acceleration rates and unseen orientations for all considered metrics in simulated data (less than 4% reduction in structural similarity index compared to similar acceleration and orientation-specific networks). The proposed reconstruction was demonstrated interactively, successfully depicting catheters in vivo with low latency (39 ms, including 19 ms for deep artifact suppression) and an image quality comparing favorably to other reconstructions. CONCLUSION: Deep artifact suppression was successfully demonstrated in the time-critical application of non-Cartesian real-time interventional cardiac MR.

Machine learning in Magnetic Resonance Imaging: Image reconstruction.

Magnetic Resonance Imaging (MRI) plays a vital role in diagnosis, management and monitoring of many diseases. However, it is an inherently slow imaging technique. Over the last 20 years, parallel imaging, temporal encoding and compressed sensing have enabled substantial speed-ups in the acquisition of MRI data, by accurately recovering missing lines of k-space data. However, clinical uptake of vastly accelerated acquisitions has been limited, in particular in compressed sensing, due to the time-consuming nature of the reconstructions and unnatural looking images. Following the success of machine learning in a wide range of imaging tasks, there has been a recent explosion in the use of machine learning in the field of MRI image reconstruction. A wide range of approaches have been proposed, which can be applied in k-space and/or image-space. Promising results have been demonstrated from a range of methods, enabling natural looking images and rapid computation. In this review article we summarize the current machine learning approaches used in MRI reconstruction, discuss their drawbacks, clinical applications, and current trends.

Reducing Contrast Agent Dose in Cardiovascular MR Angiography with Deep Learning.

BACKGROUND: Contrast-enhanced magnetic resonance angiography (MRA) is used to assess various cardiovascular conditions. However, gadolinium-based contrast agents (GBCAs) carry a risk of dose-related adverse effects. PURPOSE: To develop a deep learning method to reduce GBCA dose by 80%. STUDY TYPE: Retrospective and prospective. POPULATION: A total of 1157 retrospective and 40 prospective congenital heart disease patients for training/validation and testing, respectively. FIELD STRENGTH/SEQUENCE: A 1.5 T, T1-weighted three-dimensional (3D) gradient echo. ASSESSMENT: A neural network was trained to enhance low-dose (LD) 3D MRA using retrospective synthetic data and tested with prospective LD data. Image quality for LD (LD-MRA), enhanced LD (ELD-MRA), and high-dose (HD-MRA) was assessed in terms of signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and a quantitative measure of edge sharpness and scored for perceptual sharpness and contrast on a 1-5 scale. Diagnostic confidence was assessed on a 1-3 scale. LD- and ELD-MRA were assessed against HD-MRA for sensitivity/specificity and agreement of vessel diameter measurements (aorta and pulmonary arteries). STATISTICAL TESTS: SNR, CNR, edge sharpness, and vessel diameters were compared between LD-, ELD-, and HD-MRA using one-way repeated measures analysis of variance with post-hoc t-tests. Perceptual quality and diagnostic confidence were compared using Friedman's test with post-hoc Wilcoxon signed-rank tests. Sensitivity/specificity was compared using McNemar's test. Agreement of vessel diameters was assessed using Bland-Altman analysis. RESULTS: SNR, CNR, edge sharpness, perceptual sharpness, and perceptual contrast were lower (P < 0.05) for LD-MRA compared to ELD-MRA and HD-MRA. SNR, CNR, edge sharpness, and perceptual contrast were comparable between ELD and HD-MRA, but perceptual sharpness was significantly lower. Sensitivity/specificity was 0.824/0.921 for LD-MRA and 0.882/0.960 for ELD-MRA. Diagnostic confidence was 2.72, 2.85, and 2.92 for LD, ELD, and HD-MRA, respectively (PLD-ELD , PLD-HD  < 0.05). Vessel diameter measurements were comparable, with biases of 0.238 (LD-MRA) and 0.278 mm (ELD-MRA). DATA CONCLUSION: Deep learning can improve contrast in LD cardiovascular MRA. LEVEL OF EVIDENCE LEVEL: 2 TECHNICAL EFFICACY: Stage 2.

Primary role for melatonin MT2 receptors in the regulation of anhedonia and circadian temperature rhythm.

Circadian rhythms disturbance is widely observable in patients with major depression (MD) and is also associated with depression vulnerability. Of them, disturbed melatonin secretion rhythm is particularly relevant to MD and is strongly phase-locked to core body temperature (CBT) rhythm. Here we aim to study the specific role of each melatonin receptor (MT1 and MT2) subtype in melatonin regulation of circadian CBT and its possible relationship with depressive-like behaviors. MT1-/- , MT2-/- and WT (C57BL/6) mice were used.  Anhedonia, using the sucrose intake test, circadian CBT, environmental place preference (EPP) conditioning and vulnerability to chronic social defeat stress (CSDS) procedure were studied. Moreover, the antidepressant effects of reboxetine (15 mg/kg/day, i.p.) for three weeks or ketamine (15 mg/kg i.p. every four days, 4 doses in total) were studied. Further, exposure to ultra-mild stress induced by individual housing for several weeks was also studied in these mice. MT2-/- mice showed anhedonia and lower CBT compared to WT and MT1-/-. In addition, while reward exposure raised nocturnal CBT in WT this increase did not take place in MT2-/- mice. Further, MT2-/- mice showed an enhanced vulnerability to stress-induced anhedonia and social avoidance as well as an impaired acquisition of novelty seeking behavior. Both reboxetine and ketamine reverted anhedonia and induced a clear anti-helpless behavior in the tail suspension test (TST). Reboxetine raised CBT in mice and reverted ultra-mild stress-induced anhedonia. Our findings show a primary role for MT2 receptors in the regulation of circadian CBT as well as anhedonia and suggest that these receptors could be involved in depressive disorders associated to disturbed melatonin function.

Rapid 3D whole-heart cine imaging using golden ratio stack of spirals.

Three-dimensional cine imaging provides a wealth of information about cardiac anatomy and function, but its use in the clinical environment is limited because data acquisition is very time consuming. In this work, a free-breathing 3D whole-heart cine imaging framework was developed using a time-efficient stack of spirals trajectory and accelerated reconstruction. Two suitable view ordering methods are considered with different spacing between k-space readouts in the partition dimension: uniform and tiny golden ratio based. A simulation study suggested the latter did not present any benefits in terms of similarity to the true image. The proposed method was subsequently tested on 10 prospective subjects and compared with conventional multi-slice breath-hold imaging. Image quality was evaluated using objective and subjective scores and ventricular measurements were compared to assess clinical accuracy. Image quality was lower in the proposed technique than in breath-hold images but good agreement was found in clinically relevant ventricular measurements. In addition, the proposed method was fast to acquire, required minimal planning and provided full anatomical coverage with isotropic resolution.

Golden ratio stack of spirals for flexible angiographic imaging: Proof of concept in congenital heart disease.

PURPOSE: In this study, a golden ratio stack of spiral (GRASS) sequence that used both golden step and golden angle ordering was implemented. The aim was to demonstrate that GRASS acquisitions could be flexibly reconstructed as both cardiac-gated and time-resolved angiograms. METHODS: Image quality of time-resolved and cardiac-gated reconstructions of the GRASS sequence were compared to 3 conventional stack of spirals (SoS) acquisitions in an in silico model. In 10 patients, the GRASS sequence was compared to conventional breath hold angiography (BH-MRA) in terms of image quality and for vessel measurement. Vessel measurements were also compared to cine images. RESULTS: In the cardiac-gated in silico model, the image quality of GRASS was superior to regular and golden-angle with regular step SoS approaches. In the time-resolved model, GRASS image quality was comparable to the golden-angle with regular step technique and superior to regular SoS acquisitions. In patients, there was no difference in qualitative image scores between GRASS and BH-MRA, but SNR was lower. There was good agreement in vessel measurements between the GRASS reconstructions and conventional MR techniques (BH-MRA: 29.8 ± 5.6 mm, time-resolved GRASS-MRA: 29.9 ± 5.4 mm, SSFP diastolic: 29.4 ± 5.8 mm, cardiac-gated GRASS-MRA diastolic: 29.5 ± 5.5 mm, P > 0.87). CONCLUSION: We have demonstrated that the GRASS acquisition enables flexible reconstruction of the same raw data as both time-resolved and cardiac-gated volumes. This may enable better interrogation of anatomy in congenital heart disease.