Temporal phase unwrapping for CINE velocity imaging

A comparison of phase unwrapping methods in velocity-encoded MRI for aortic flows

PURPOSE

The phase of a MRI signal is used to encode the velocity of blood flow. Phase unwrapping artifacts may appear when aiming to improve the velocity-to-noise ratio (VNR) of the measured velocity field. This study aims to compare various unwrapping algorithms on ground-truth synthetic data generated using computational fluid dynamics (CFD) simulations.

METHODS

We compare four different phase unwrapping algorithms on two different synthetic datasets of four-dimensional flow MRI and 26 datasets of 2D PC-MRI acquisitions including the ascending and descending aorta. The synthetic datasets are constructed using CFD simulations of an aorta with a coarctation, with different levels of spatiotemporal resolutions and noise. The error of the unwrapped images was assessed by comparison against the ground truth velocity field in the synthetic data and dual-VENC reconstructions in the in vivo data.

RESULTS

Using the unwrapping algorithms, we were able to remove aliased voxels in the data almost entirely, reducing the L2-error compared to the ground truth by 50%-80%. Results indicated that the best choice of algorithm depend on the spatiotemporal resolution and noise level of the dataset. Temporal unwrapping is most successful with a high temporal and low spatial resolution (δ t = 30 $$ \delta t=30 $$  ms,h = 2 . 5 $$ h=2.5 $$  mm), reducing the L2-error by 70%-85%, while Laplacian unwrapping performs better with a lower temporal or better spatial resolution (δ t = 60 $$ \delta t=60 $$ ms,h = 1 . 5 $$ h=1.5 $$ mm), especially for signal-to-noise ratio (SNR) 12 as opposed to SNR 15, with an error reduction of 55%-85% compared to the 50%-75% achieved by the Temporal method. The differences in performance between the methods are statistically significant.

CONCLUSIONS

The temporal method and spatiotemporal Laplacian method provide the best results, with the spatiotemporal Laplacian being more robust. However, single-V enc $$ {V}_{\mathrm{enc}} $$ methods only situationally and not generally reach the performance of dual-V enc $$ {V}_{\mathrm{enc}} $$ unwrapping methods.

Four-Dimensional Flow MR Imaging: Technique and Advances

4D Flow MRI is an advanced imaging technique for comprehensive non-invasive assessment of the cardiovascular system. The capture of the blood velocity vector field throughout the cardiac cycle enables measures of flow, pulse wave velocity, kinetic energy, wall shear stress, and more. Advances in hardware, MRI data acquisition and reconstruction methodology allow for clinically feasible scan times. The availability of 4D Flow analysis packages allows for more widespread use in research and the clinic and will facilitate much needed multi-center, multi-vendor studies in order to establish consistency across scanner platforms and to enable larger scale studies to demonstrate clinical value.

Automatic Time-Resolved Cardiovascular Segmentation of 4D Flow MRI Using Deep Learning

BACKGROUND

Segmenting the whole heart over the cardiac cycle in 4D flow MRI is a challenging and time-consuming process, as there is considerable motion and limited contrast between blood and tissue.

PURPOSE

To develop and evaluate a deep learning-based segmentation method to automatically segment the cardiac chambers and great thoracic vessels from 4D flow MRI.

STUDY TYPE

Retrospective.

SUBJECTS

A total of 205 subjects, including 40 healthy volunteers and 165 patients with a variety of cardiac disorders were included. Data were randomly divided into training (n = 144), validation (n = 20), and testing (n = 41) sets.

FIELD STRENGTH/SEQUENCE

A 3 T/time-resolved velocity encoded 3D gradient echo sequence (4D flow MRI).

ASSESSMENT

A 3D neural network based on the U-net architecture was trained to segment the four cardiac chambers, aorta, and pulmonary artery. The segmentations generated were compared to manually corrected atlas-based segmentations. End-diastolic (ED) and end-systolic (ES) volumes of the four cardiac chambers were calculated for both segmentations.

STATISTICAL TESTS

Dice score, Hausdorff distance, average surface distance, sensitivity, precision, and miss rate were used to measure segmentation accuracy. Bland-Altman analysis was used to evaluate agreement between volumetric parameters.

RESULTS

The following evaluation metrics were computed: mean Dice score (0.908 ± 0.023) (mean ± SD), Hausdorff distance (1.253 ± 0.293 mm), average surface distance (0.466 ± 0.136 mm), sensitivity (0.907 ± 0.032), precision (0.913 ± 0.028), and miss rate (0.093 ± 0.032). Bland-Altman analyses showed good agreement between volumetric parameters for all chambers. Limits of agreement as percentage of mean chamber volume (LoA%), left ventricular: 9.3%, 13.5%, left atrial: 12.4%, 16.9%, right ventricular: 9.9%, 15.6%, and right atrial: 18.7%, 14.4%; for ED and ES, respectively.

DATA CONCLUSION

The addition of this technique to the 4D flow MRI assessment pipeline could expedite and improve the utility of this type of acquisition in the clinical setting.

EVIDENCE LEVEL

4 TECHNICAL EFFICACY: Stage 1.

Advances in machine learning applications for cardiovascular 4D flow MRI

Four-dimensional flow magnetic resonance imaging (MRI) has evolved as a non-invasive imaging technique to visualize and quantify blood flow in the heart and vessels. Hemodynamic parameters derived from 4D flow MRI, such as net flow and peak velocities, but also kinetic energy, turbulent kinetic energy, viscous energy loss, and wall shear stress have shown to be of diagnostic relevance for cardiovascular diseases. 4D flow MRI, however, has several limitations. Its long acquisition times and its limited spatio-temporal resolutions lead to inaccuracies in velocity measurements in small and low-flow vessels and near the vessel wall. Additionally, 4D flow MRI requires long post-processing times, since inaccuracies due to the measurement process need to be corrected for and parameter quantification requires 2D and 3D contour drawing. Several machine learning (ML) techniques have been proposed to overcome these limitations. Existing scan acceleration methods have been extended using ML for image reconstruction and ML based super-resolution methods have been used to assimilate high-resolution computational fluid dynamic simulations and 4D flow MRI, which leads to more realistic velocity results. ML efforts have also focused on the automation of other post-processing steps, by learning phase corrections and anti-aliasing. To automate contour drawing and 3D segmentation, networks such as the U-Net have been widely applied. This review summarizes the latest ML advances in 4D flow MRI with a focus on technical aspects and applications. It is divided into the current status of fast and accurate 4D flow MRI data generation, ML based post-processing tools for phase correction and vessel delineation and the statistical evaluation of blood flow.

Fetal cardiovascular blood flow MRI: techniques and applications

Fetal cardiac MRI is challenging due to fetal and maternal movements as well as the need for a reliable cardiac gating signal and high spatiotemporal resolution. Ongoing research and recent technical developments to address these challenges show the potential of MRI as an adjunct to ultrasound for the assessment of the fetal heart and great vessels. MRI measurements of blood flow have enabled the assessment of normal fetal circulation as well as conditions with disrupted circulations, such as congenital heart disease, along with associated organ underdevelopment and hemodynamic instability. This review provides details of the techniques used in fetal cardiovascular blood flow MRI, including single slice and volumetric imaging sequences, post-processing and analysis, along with a summary of applications in human studies and animal models.

Deep learning-based velocity antialiasing of 4D-flow MRI

Purpose

To develop a convolutional neural network (CNN) for the robust and fast correction of velocity aliasing in 4D‐flow MRI.

Methods

This study included 667 adult subjects with aortic 4D‐flow MRI data with existing velocity aliasing (n = 362) and no velocity aliasing (n = 305). Additionally, 10 controls received back‐to‐back 4D‐flow scans with systemically varied velocity‐encoding sensitivity (vencs) at 60, 100, and 175 cm/s. The no‐aliasing data sets were used to simulate velocity aliasing by reducing the venc to 40%–70% of the original, alongside a ground truth locating all aliased voxels (153 training, 152 testing). The 152 simulated and 362 existing aliasing data sets were used for testing and compared with a conventional velocity antialiasing algorithm. Dice scores were calculated to quantify CNN performance. For controls, the venc 175‐cm/s scans were used as the ground truth and compared with the CNN‐corrected venc 60 and 100 cm/s data sets

Results

The CNN required 176 ± 30 s to perform compared with 162 ± 14 s for the conventional algorithm. The CNN showed excellent performance for the simulated data compared with the conventional algorithm (median range of Dice scores CNN: [0.89–0.99], conventional algorithm: [0.84–0.94], p < 0.001, across all simulated vencs) and detected more aliased voxels in existing velocity aliasing data sets (median detected CNN: 159 voxels [31–605], conventional algorithm: 65 [7–417], p < 0.001). For controls, the CNN showed Dice scores of 0.98 [0.95–0.99] and 0.96 [0.87–0.99] for venc = 60 cm/s and 100 cm/s, respectively, while flow comparisons showed moderate‐excellent agreement.

Conclusion

Deep learning enabled fast and robust velocity anti‐aliasing in 4D‐flow MRI.

Measuring Arterial Pulsatility With Dynamic Inflow Magnitude Contrast

The pulsatility of blood flow through cerebral arteries is clinically important, as it is intrinsically associated with cerebrovascular health. In this study we outline a new MRI approach to measuring the real-time pulsatile flow in cerebral arteries, which is based on the inflow phenomenon associated with fast gradient-recalled-echo acquisitions. Unlike traditional phase-contrast techniques, this new method, which we dub dynamic inflow magnitude contrast (DIMAC), does not require velocity-encoding gradients as sensitivity to flow velocity is derived purely from the inflow effect. We achieved this using a highly accelerated single slice EPI acquisition with a very short TR (15 ms) and a 90° flip angle, thus maximizing inflow contrast. We simulate the spoiled GRE signal in the presence of large arteries and perform a sensitivity analysis. The sensitivity analysis demonstrates that in the regime of high inflow contrast, DIMAC shows much greater sensitivity to flow velocity over blood volume changes. We support this theoretical prediction with in-vivo data collected in two separate experiments designed to demonstrate the utility of the DIMAC signal contrast. We perform a hypercapnia challenge experiment in order to experimentally modulate arterial tone within subjects, and thus modulate the arterial pulsatile flow waveform. We also perform a thigh-cuff release challenge, designed to induce a transient drop in blood pressure, and demonstrate that the continuous DIMAC signal captures the complex transient change in the pulsatile and non-pulsatile components of flow. In summary, this study proposes a new role for a well-established source of MR image contrast and demonstrates its potential for measuring both steady-state and dynamic changes in arterial tone.

Divergence-Free Constrained Phase Unwrapping and Denoising for 4D Flow MRI Using Weighted Least-Squares

A novel divergence-free constrained phase unwrapping method was proposed and evaluated for 4D flow MRI. The unwrapped phase field was obtained by integrating the phase variations estimated from the wrapped phase data using weighted least-squares. The divergence-free constraint for incompressible blood flow was incorporated to regulate and denoise the resulting phase field. The proposed method was tested on synthetic phase data of left ventricular flow and in vitro 4D flow measurement of Poiseuille flow. The method was additionally applied to in vivo 4D flow measurements in the thoracic aorta from 30 human subjects. The performance of the proposed method was compared to the state-of-the-art 4D single-step Laplacian algorithm. The synthetic phase data were completely unwrapped by the proposed method for all the cases with velocity encoding (venc) as low as 20% of the maximum velocity and signal-to-noise ratio as low as 5. The in vitro Poiseuille flow data were completely unwrapped with a 60% increase in the velocity-to-noise ratio. For the in-vivo aortic datasets with venc ratio less than 0.4, the proposed method significantly improved the success rate by as much as 40% and reduced the velocity error levels by a factor of 10 compared to the state-of-the-art method. The divergence-free constrained method exhibits reliability and robustness on phase unwrapping and shows improved accuracy of velocity and hemodynamic quantities by unwrapping the low-venc 4D flow MRI data.

Non-invasive Assessment of Systolic and Diastolic Cardiac Function During Rest and Stress Conditions Using an Integrated Image-Modeling Approach

Background: The possibility of non-invasively assessing load-independent parameters characterizing cardiac function is of high clinical value. Typically, these parameters are assessed during resting conditions. However, for diagnostic purposes, the parameter behavior across a physiologically relevant range of heart rate and loads is more relevant than the isolated measurements performed at rest. This study sought to evaluate changes in non-invasive estimations of load-independent parameters of left-ventricular contraction and relaxation patterns at rest and during dobutamine stress. Methods: We applied a previously developed approach that combines non-invasive measurements with a physiologically-based, reduced-order model of the cardiovascular system to provide subject-specific estimates of parameters characterizing left ventricular function. In this model, the contractile state of the heart at each time point along the cardiac cycle is modeled using a time-varying elastance curve. Non-invasive data, including four-dimensional magnetic resonance imaging (4D Flow MRI) measurements, were acquired in nine subjects without a known heart disease at rest and during dobutamine stress. For each of the study subjects, we constructed two personalized models corresponding to the resting and the stress state. Results: Applying the modeling framework, we identified significant increases in the left ventricular contraction rate constant [from 1.5 ± 0.3 to 2 ± 0.5 (p = 0.038)] and relaxation constant [from 37.2 ± 6.9 to 46.1 ± 12 (p = 0.028)]. In addition, we found a significant decrease in the elastance diastolic time constant from 0.4 ± 0.04 s to 0.3 ± 0.03 s (p = 0.008). Conclusions: The integrated image-modeling approach allows the assessment of cardiovascular function given as model-based parameters. The agreement between the estimated parameter values and previously reported effects of dobutamine demonstrates the potential of the approach to assess advanced metrics of pathophysiology that are otherwise difficult to obtain non-invasively in clinical practice.

Age-related changes of shape and flow dynamics in healthy adult aortas: A 4D flow MRI study

BACKGROUND

Abnormal flow dynamics play an early and causative role in pathologic changes of the ascending aorta.

PURPOSE

To identify: 1) the changes in flow, shape, and size that occur in the ascending aorta with normal human ageing and 2) the influence of these factors on aortic flow dynamics.

STUDY TYPE

Retrospective.

SUBJECTS

In all, 247 subjects (age range 19-86 years, mean 49 ± 17.7, 169 males) free of aortic or aortic valve pathology were included in this study. Subjects were stratified by youngest (18-33 years; n = 64), highest (>60 years, n = 67), and the middle two quartiles (34-60 years, n = 116).

FIELD STRENGTH/SEQUENCE

Subjects underwent a cardiac MRI (3T) exam including 4D-flow MRI of the aorta.

ASSESSMENT

Aortic curvature, arch shape, ascending aortic angle, ascending aortic diameter, and the stroke volume normalized by the aortic volume (nSV) were measured. Velocity, vorticity, and helicity were quantified across the thoracic aorta.

STATISTICAL TESTS

Univariate and multivariate regressions were used to quantify continuous relationships between variables.

RESULTS

Aortic diameter, ascending aortic angle, shape, and curvature all increased across age while nSV decreased (all P < 0.0001). Systolic vorticity in the mid arch decreased by 50% across the age range (P < 0.0001), while peak helicity decreased by 80% (P < 0.0001). Curvature tightly governs optimal flow in the youngest quartile, with an effect size 1.5 to 4 times larger than other parameters in the descending aorta, but had a minimal influence with advancing age. In the upper quartile of age, flow dynamics were almost completely determined by nSV, exerting an effect size on velocity and vorticity >10 times that of diameter and other shape factors.

DATA CONCLUSION

Aortic shape influences flow dynamics in younger subjects. Flow conditions become increasingly disturbed with advancing age, and in these conditions nSV has a more dominant effect on flow patterns than shape factors.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;49:90-100.

Left Atrial 4D Blood Flow Dynamics and Hemostasis following Electrical Cardioversion of Atrial Fibrillation

Background: Electrical cardioversion in patients with atrial fibrillation is followed by a transiently impaired atrial mechanical function, termed atrial stunning. During atrial stunning, a retained risk of left atrial thrombus formation exists, which may be attributed to abnormal left atrial blood flow patterns. 4D Flow cardiovascular magnetic resonance (CMR) enables blood flow assessment from the entire three-dimensional atrial volume throughout the cardiac cycle. We sought to investigate left atrial 4D blood flow patterns and hemostasis during left atrial stunning and after left atrial mechanical function was restored.

Methods: 4D Flow and morphological CMR data as well as blood samples were collected in fourteen patients at two time-points: 2–3 h (Time-1) and 4 weeks (Time-2) following cardioversion. The volume of blood stasis and duration of blood stasis were calculated. In addition, hemostasis markers were analyzed.

Results: From Time-1 to Time-2: Heart rate decreased (61 ± 7 vs. 56 ± 8 bpm, p = 0.01); Maximum change in left atrial volume increased (8 ± 4 vs. 22 ± 15%, p = 0.009); The duration of stasis (68 ± 11 vs. 57 ± 8%, p = 0.002) and the volume of stasis (14 ± 9 vs. 9 ± 7%, p = 0.04) decreased; Thrombin-antithrombin complex (TAT) decreased (5.2 ± 3.3 vs. 3.3 ± 2.2 μg/L, p = 0.008). A significant correlation was found between TAT and the volume of stasis (r² = 0.69, p < 0.001) at Time-1 and between TAT and the duration of stasis (r² = 0.34, p = 0.04) at Time-2.

Conclusion: In this longitudinal study, left atrial multidimensional blood flow was altered and blood stasis was elevated during left atrial stunning compared to the restored left atrial mechanical function. The coagulability of blood was also elevated during atrial stunning. The association between blood stasis and hypercoagulability proposes that assessment of left atrial 4D flow can add to the pathophysiological understanding of thrombus formation during atrial fibrillation related atrial stunning.

Mechanical dyssynchrony alters left ventricular flow energetics in failing hearts with LBBB: a 4D flow CMR pilot study

The impact of left bundle branch block (LBBB) related mechanical dyssynchrony on left ventricular (LV) diastolic function remains unclear. 4D flow cardiovascular magnetic resonance (CMR) has provided reliable markers of LV dysfunction: reduced volume and kinetic energy (KE) of the portion of LV inflow which passes directly to outflow (Direct Flow) has been demonstrated in failing hearts compared to normal hearts. We sought to investigate the impact of mechanical dyssynchrony on diastolic function by comparing 4D flow in myopathic LVs with and without LBBB. CMR data were acquired at 3 T in 22 heart failure patients; 11 with LBBB and 11 without LBBB matched according to several demographic and clinical parameters. An established 4D flow analysis method was used to separate the LV end-diastolic (ED) volume into functional flow components based on the blood’s timing and route through the heart cavities. While the Direct Flow volume was not different between the groups, the KE possessed at ED was lower in LBBB patients (P = 0.018). Direct Flow entering the LV during early diastolic filling possessed less KE at ED in LBBB patients compared to non-LBBB patients, whereas no intergroup difference was observed during late filling. Pre-systolic KE of LV Direct Flow was reduced in patients with LBBB compared to matched patients with normal conduction. These intriguing findings propose that 4D flow specific measures can serve as markers of LV mechanical dyssynchrony in heart failure patients, and could possibly be investigated as predictors of response to cardiac resynchronization therapy.

Bridging the gap between measurements and modelling: a cardiovascular functional avatar

Lumped parameter models of the cardiovascular system have the potential to assist researchers and clinicians to better understand cardiovascular function. The value of such models increases when they are subject specific. However, most approaches to personalize lumped parameter models have thus far required invasive measurements or fall short of being subject specific due to a lack of the necessary clinical data. Here, we propose an approach to personalize parameters in a model of the heart and the systemic circulation using exclusively non-invasive measurements. The personalized model is created using flow data from four-dimensional magnetic resonance imaging and cuff pressure measurements in the brachial artery. We term this personalized model the cardiovascular avatar. In our proof-of-concept study, we evaluated the capability of the avatar to reproduce pressures and flows in a group of eight healthy subjects. Both quantitatively and qualitatively, the model-based results agreed well with the pressure and flow measurements obtained in vivo for each subject. This non-invasive and personalized approach can synthesize medical data into clinically relevant indicators of cardiovascular function, and estimate hemodynamic variables that cannot be assessed directly from clinical measurements.

Improving visualization of 4D flow cardiovascular magnetic resonance with four-dimensional angiographic data: generation of a 4D phase-contrast magnetic resonance CardioAngiography (4D PC-MRCA)

Magnetic Resonance Angiography (MRA) and Phase-Contrast MRA (PC-MRA) approaches used for assessment of cardiovascular morphology typically result in data containing information from the entire cardiac cycle combined into one 2D or 3D image. Information specific to each timeframe of the cardiac cycle is, however, lost in this process. This study proposes a novel technique, called Phase-Contrast Magnetic Resonance CardioAngiography (4D PC-MRCA), that utilizes the full potential of 4D Flow CMR when generating temporally resolved PC-MRA data to improve visualization of the heart and major vessels throughout the cardiac cycle. Using non-rigid registration between the timeframes of the 4D Flow CMR acquisition, the technique concentrates information from the entire cardiac cycle into an angiographic dataset at one specific timeframe, taking movement over the cardiac cycle into account. Registration between the timeframes is used once more to generate a time-resolved angiography. The method was evaluated in ten healthy volunteers. Visual comparison of the 4D PC-MRCAs versus PC-MRAs generated from 4D Flow CMR using the traditional approach was performed by two observers using Maximum Intensity Projections (MIPs). The 4D PC-MRCAs resulted in better visibility of the main anatomical regions of the cardiovascular system, especially where cardiac or vessel motion was present. The proposed method represents an improvement over previous PC-MRA generation techniques that rely on 4D Flow CMR, as it effectively utilizes all the information available in the acquisition. The 4D PC-MRCA can be used to visualize the motion of the heart and major vessels throughout the entire cardiac cycle.

Left ventricular hemodynamic forces as a marker of mechanical dyssynchrony in heart failure patients with left bundle branch block

Left bundle branch block (LBBB) causes left ventricular (LV) dyssynchrony which is often associated with heart failure. A significant proportion of heart failure patients do not demonstrate clinical improvement despite cardiac resynchronization therapy (CRT). How LBBB-related effects on LV diastolic function may contribute to those therapeutic failures has not been clarified. We hypothesized that LV hemodynamic forces calculated from 4D flow MRI could serve as a marker of diastolic mechanical dyssynchrony in LBBB hearts. MRI data were acquired in heart failure patients with LBBB or matched patients without LBBB. LV pressure gradients were calculated from the Navier-Stokes equations. Integration of the pressure gradients over the LV volume rendered the hemodynamic forces. The findings demonstrate that the LV filling forces are more orthogonal to the main LV flow direction in heart failure patients with LBBB compared to those without LBBB during early but not late diastole. The greater the conduction abnormality the greater the discordance of LV filling force with the predominant LV flow direction (r² = 0.49). Such unique flow-specific measures of mechanical dyssynchrony may serve as an additional tool for considering the risks imposed by conduction abnormalities in heart failure patients and prove to be useful in predicting response to CRT.

Overview of quantitative susceptibility mapping

Magnetic susceptibility describes the magnetizability of a material to an applied magnetic field and represents an important parameter in the field of MRI. With the recently introduced method of quantitative susceptibility mapping (QSM) and its conceptual extension to susceptibility tensor imaging (STI), the non-invasive assessment of this important physical quantity has become possible with MRI. Both methods solve the ill-posed inverse problem to determine the magnetic susceptibility from local magnetic fields. Whilst QSM allows the extraction of the spatial distribution of the bulk magnetic susceptibility from a single measurement, STI enables the quantification of magnetic susceptibility anisotropy, but requires multiple measurements with different orientations of the object relative to the main static magnetic field. In this review, we briefly recapitulate the fundamental theoretical foundation of QSM and STI, as well as computational strategies for the characterization of magnetic susceptibility with MRI phase data. In the second part, we provide an overview of current methodological and clinical applications of QSM with a focus on brain imaging. Copyright © 2016 John Wiley & Sons, Ltd.

Improving left ventricular segmentation in four-dimensional flow MRI using intramodality image registration for cardiac blood flow analysis

PURPOSE

Assessment of blood flow in the left ventricle using four-dimensional flow MRI requires accurate left ventricle segmentation that is often hampered by the low contrast between blood and the myocardium. The purpose of this work is to improve left-ventricular segmentation in four-dimensional flow MRI for reliable blood flow analysis.

METHOD

The left ventricle segmentations are first obtained using morphological cine-MRI with better in-plane resolution and contrast, and then aligned to four-dimensional flow MRI data. This alignment is, however, not trivial due to inter-slice misalignment errors caused by patient motion and respiratory drift during breath-hold based cine-MRI acquisition. A robust image registration based framework is proposed to mitigate such errors automatically. Data from 20 subjects, including healthy volunteers and patients, was used to evaluate its geometric accuracy and impact on blood flow analysis.

RESULTS

High spatial correspondence was observed between manually and automatically aligned segmentations, and the improvements in alignment compared to uncorrected segmentations were significant (P < 0.01). Blood flow analysis from manual and automatically corrected segmentations did not differ significantly (P > 0.05).

CONCLUSION

Our results demonstrate the efficacy of the proposed approach in improving left-ventricular segmentation in four-dimensional flow MRI, and its potential for reliable blood flow analysis. Magn Reson Med 79:554-560, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Creating hemodynamic atlases of cardiac 4D flow MRI

Purpose

Hemodynamic atlases can add to the pathophysiological understanding of cardiac diseases. This study proposes a method to create hemodynamic atlases using 4D Flow magnetic resonance imaging (MRI). The method is demonstrated for kinetic energy (KE) and helicity density (H_d).

Materials and Methods

Thirteen healthy subjects underwent 4D Flow MRI at 3T. Phase‐contrast magnetic resonance cardioangiographies (PC‐MRCAs) and an average heart were created and segmented. The PC‐MRCAs, KE, and H_d were nonrigidly registered to the average heart to create atlases. The method was compared with 1) rigid, 2) affine registration of the PC‐MRCAs, and 3) affine registration of segmentations. The peak and mean KE and H_d before and after registration were calculated to evaluate interpolation error due to nonrigid registration.

Results

The segmentations deformed using nonrigid registration overlapped (median: 92.3%) more than rigid (23.1%, P < 0.001), and affine registration of PC‐MRCAs (38.5%, P < 0.001) and affine registration of segmentations (61.5%, P < 0.001). The peak KE was 4.9 mJ using the proposed method and affine registration of segmentations (P = 0.91), 3.5 mJ using rigid registration (P < 0.001), and 4.2 mJ using affine registration of the PC‐MRCAs (P < 0.001). The mean KE was 1.1 mJ using the proposed method, 0.8 mJ using rigid registration (P < 0.001), 0.9 mJ using affine registration of the PC‐MRCAs (P < 0.001), and 1.0 mJ using affine registration of segmentations (P = 0.028). The interpolation error was 5.2 ± 2.6% at mid‐systole, 2.8 ± 3.8% at early diastole for peak KE; 9.6 ± 9.3% at mid‐systole, 4.0 ± 4.6% at early diastole, and 4.9 ± 4.6% at late diastole for peak H_d. The mean KE and H_d were not affected by interpolation.

Conclusion

Hemodynamic atlases can be obtained with minimal user interaction using nonrigid registration of 4D Flow MRI.

Level of Evidence: 2

Technical Efficacy: Stage 1

J. Magn. Reson. Imaging 2017;46:1389–1399.

Assessment of left ventricular hemodynamic forces in healthy subjects and patients with dilated cardiomyopathy using 4D flow MRI

We hypothesized that the direction of global left ventricular (LV) hemodynamic forces during diastolic filling are concordant with the main flow axes in normal LVs, but that this pattern would be altered in dilated and dysfunctional LVs. Therefore, we aimed to assess the LV hemodynamic filling forces in a group of healthy subjects and compare them to the results from a group of patients with dilated cardiomyopathy (DCM). Ten healthy subjects and 10 DCM patients were enrolled. Morphological short‐ (SAx) and long‐axis (LAx) images and 4D flow MRI data were acquired at 1.5T. The LV pressure gradients were computed from the 4D flow data using the Navier–Stokes equations. By integrating the pressure gradients over the LV volume at each time frame, the magnitude and direction of the global hemodynamic force was calculated over the cardiac cycle. The hemodynamic forces acting in the SAx‐ and LAx‐directions were used to calculate the “SAx‐max/LAx‐max”‐ratio for the early (E‐wave) and late (A‐wave) diastolic filling. In the LAx‐plane, the temporal progression of the hemodynamic force followed a consistent pattern in the healthy subjects. The “SAx‐max/LAx‐max”‐ratio was significantly larger at both E‐wave (0.53 ± 0.15 vs. 0.23 ± 0.12, P < 0.0001) and A‐wave (0.44 ± 0.21 vs. 0.26 ± 0.09, P < 0.03) in the DCM patients compared to the healthy subjects. 4D flow MRI data allow quantification of LV hemodynamic forces acting on the LV myocardial wall. The LV hemodynamic filling forces showed a similar temporal progression among healthy subjects, whereas DCM patients had forces that were more heterogeneous in their direction and magnitude during diastole.

Advanced Analysis Techniques for Intra-cardiac Flow Evaluation from 4D Flow MRI

PURPOSE OF THE REVIEW

Time-resolved 3D velocity-encoded MR imaging with velocity encoding in three directions (4D Flow) has emerged as a novel MR acquisition technique providing detailed information on flow in the cardiovascular system. In contrast to other clinically available imaging techniques such as echo-Doppler, 4D Flow MRI provides the 3D Flow velocity field within a volumetric region of interest over the cardiac cycle. This work reviews the most recent advances in the development and application of dedicated image analysis techniques for the assessment of intra-cardiac flow features from 4D Flow MRI.

RECENT FINDINGS

Novel image analysis techniques have been developed for extraction of relevant intra-cardiac flow features from 4D Flow MRI, which have been successfully applied in various patient cohorts and volunteer studies. Disturbed flow patterns have been linked with valvular abnormalities and ventricular dysfunction. Recent technical advances have resulted in reduced scan times and improvements in image quality, increasing the potential clinical applicability of 4D Flow MRI.

SUMMARY

4D Flow MRI provides unique capabilities for 3D visualization and quantification of intra-cardiac blood flow. Contemporary knowledge on 4D Flow MRI shows promise for further exploration of the potential use of the technique in research and clinical applications.

Atlas-based analysis of 4D flow CMR: automated vessel segmentation and flow quantification

BACKGROUND

Flow volume quantification in the great thoracic vessels is used in the assessment of several cardiovascular diseases. Clinically, it is often based on semi-automatic segmentation of a vessel throughout the cardiac cycle in 2D cine phase-contrast Cardiovascular Magnetic Resonance (CMR) images. Three-dimensional (3D), time-resolved phase-contrast CMR with three-directional velocity encoding (4D flow CMR) permits assessment of net flow volumes and flow patterns retrospectively at any location in a time-resolved 3D volume. However, analysis of these datasets can be demanding. The aim of this study is to develop and evaluate a fully automatic method for segmentation and analysis of 4D flow CMR data of the great thoracic vessels.

METHODS

The proposed method utilizes atlas-based segmentation to segment the great thoracic vessels in systole, and registration between different time frames of the cardiac cycle in order to segment these vessels over time. Additionally, net flow volumes are calculated automatically at locations of interest. The method was applied on 4D flow CMR datasets obtained from 11 healthy volunteers and 10 patients with heart failure. Evaluation of the method was performed visually, and by comparison of net flow volumes in the ascending aorta obtained automatically (using the proposed method), and semi-automatically. Further evaluation was done by comparison of net flow volumes obtained automatically at different locations in the aorta, pulmonary artery, and caval veins.

RESULTS

Visual evaluation of the generated segmentations resulted in good outcomes for all the major vessels in all but one dataset. The comparison between automatically and semi-automatically obtained net flow volumes in the ascending aorta resulted in very high correlation (r (2)=0.926). Moreover, comparison of the net flow volumes obtained automatically in other vessel locations also produced high correlations where expected: pulmonary trunk vs. proximal ascending aorta (r (2)=0.955), pulmonary trunk vs. pulmonary branches (r (2)=0.808), and pulmonary trunk vs. caval veins (r (2)=0.906).

CONCLUSIONS

The proposed method allows for automatic analysis of 4D flow CMR data, including vessel segmentation, assessment of flow volumes at locations of interest, and 4D flow visualization. This constitutes an important step towards facilitating the clinical utility of 4D flow CMR.

4D flow cardiovascular magnetic resonance consensus statement

Pulsatile blood flow through the cavities of the heart and great vessels is time-varying and multidirectional. Access to all regions, phases and directions of cardiovascular flows has formerly been limited. Four-dimensional (4D) flow cardiovascular magnetic resonance (CMR) has enabled more comprehensive access to such flows, with typical spatial resolution of 1.5×1.5×1.5 - 3×3×3 mm(3), typical temporal resolution of 30-40 ms, and acquisition times in the order of 5 to 25 min. This consensus paper is the work of physicists, physicians and biomedical engineers, active in the development and implementation of 4D Flow CMR, who have repeatedly met to share experience and ideas. The paper aims to assist understanding of acquisition and analysis methods, and their potential clinical applications with a focus on the heart and greater vessels. We describe that 4D Flow CMR can be clinically advantageous because placement of a single acquisition volume is straightforward and enables flow through any plane across it to be calculated retrospectively and with good accuracy. We also specify research and development goals that have yet to be satisfactorily achieved. Derived flow parameters, generally needing further development or validation for clinical use, include measurements of wall shear stress, pressure difference, turbulent kinetic energy, and intracardiac flow components. The dependence of measurement accuracy on acquisition parameters is considered, as are the uses of different visualization strategies for appropriate representation of time-varying multidirectional flow fields. Finally, we offer suggestions for more consistent, user-friendly implementation of 4D Flow CMR acquisition and data handling with a view to multicenter studies and more widespread adoption of the approach in routine clinical investigations.

Cardiovascular magnetic resonance phase contrast imaging

Cardiovascular magnetic resonance (CMR) phase contrast imaging has undergone a wide range of changes with the development and availability of improved calibration procedures, visualization tools, and analysis methods. This article provides a comprehensive review of the current state-of-the-art in CMR phase contrast imaging methodology, clinical applications including summaries of past clinical performance, and emerging research and clinical applications that utilize today's latest technology.

Quantitative phase microscopy: automated background leveling techniques and smart temporal phase unwrapping

In order for time-dynamic quantitative phase microscopy to yield meaningful data to scientists, raw phase measurements must be converted to sequential time series that are consistently phase unwrapped with minimal residual background shape. Beyond the initial phase unwrapping, additional steps must be taken to convert the phase to time-meaningful data sequences. This consists of two major operations both outlined in this paper and shown to operate robustly on biological datasets. An automated background leveling procedure is introduced that consistently removes background shape and minimizes mean background phase value fluctuations. By creating a background phase value that is stable over time, the phase values of features of interest can be examined as a function of time to draw biologically meaningful conclusions. Residual differences between sequential frames of data can be present due to inconsistent phase unwrapping, causing localized regions to have phase values at similar object locations inconsistently changed by large values between frames, not corresponding to physical changes in the sample being observed. This is overcome by introducing a new method, referred to as smart temporal unwrapping that temporally unwraps and filters the phase data such that small motion between frames is accounted for and phase data are unwrapped consistently between frames. The combination of these methods results in the creation of phase data that is stable over time by minimizing errors introduced within the processing of the raw data.

Spatial heterogeneity of four-dimensional relative pressure fields in the human left ventricle

PURPOSE

To assess the spatial heterogeneity of the four-dimensional (4D) relative pressure fields in the healthy human left ventricle (LV) and provide reference data for normal LV relative pressure.

METHODS

Twelve healthy subjects underwent a cardiac MRI examination where 4D flow and morphological data were acquired. The latter data were segmented and used to define the borders of the LV for computation of relative pressure fields using the pressure Poisson equation. The LV lumen was divided into 17 pie-shaped segments.

RESULTS

In the normal left ventricle, the relative pressure in the apical segments was significantly higher relative to the basal segments (P < 0.0005) along both the anteroseptal and inferolateral sides after the peaks of early (E-wave) and late (A-wave) diastolic filling. The basal anteroseptal segment showed significantly lower median pressure than the opposite basal inferolateral segment during both E-wave (P < 0.0005) and A-wave (P = 0.0024).

CONCLUSION

Relative pressure in the left ventricle is heterogeneous. During diastole, the main pressure differences in the LV occur along the basal-apical axis. However, pressure differences were also found in the short axis direction and may reflect important aspects of atrioventricular coupling. Additionally, this study provides reference data on LV pressure dynamics for a group of healthy subjects.

Quantitative Phase Microscopy: how to make phase data meaningful

The continued development of hardware and associated image processing techniques for quantitative phase microscopy has allowed superior phase data to be acquired that readily shows dynamic optical volume changes and enables particle tracking. Recent efforts have focused on tying phase data and associated metrics to cell morphology. One challenge in measuring biological objects using interferometrically obtained phase information is achieving consistent phase unwrapping and -dimensions and correct for temporal discrepanices using a temporal unwrapping procedure. The residual background shape due to mean value fluctuations and residual tilts can be removed automatically using a simple object characterization algorithm. Once the phase data are processed consistently, it is then possible to characterize biological samples such as myocytes and myoblasts in terms of their size, texture and optical volume and track those features dynamically. By observing optical volume dynamically it is possible to determine the presence of objects such as vesicles within myoblasts even when they are co-located with other objects. Quantitative phase microscopy provides a label-free mechanism to characterize living cells and their morphology in dynamic environments, however it is critical to connect the measured phase to important biological function for this measurement modality to prove useful to a broader scientific community. In order to do so, results must be highly consistent and require little to no user manipulation to achieve high quality nynerical results that can be combined with other imaging modalities.

Volume Tracking: A new method for quantitative assessment and visualization of intracardiac blood flow from three-dimensional, time-resolved, three-component magnetic resonance velocity mapping

Background

Functional and morphological changes of the heart influence blood flow patterns. Therefore, flow patterns may carry diagnostic and prognostic information. Three-dimensional, time-resolved, three-directional phase contrast cardiovascular magnetic resonance (4D PC-CMR) can image flow patterns with unique detail, and using new flow visualization methods may lead to new insights. The aim of this study is to present and validate a novel visualization method with a quantitative potential for blood flow from 4D PC-CMR, called Volume Tracking, and investigate if Volume Tracking complements particle tracing, the most common visualization method used today.

Methods

Eight healthy volunteers and one patient with a large apical left ventricular aneurysm underwent 4D PC-CMR flow imaging of the whole heart. Volume Tracking and particle tracing visualizations were compared visually side-by-side in a visualization software package. To validate Volume Tracking, the number of particle traces that agreed with the Volume Tracking visualizations was counted and expressed as a percentage of total released particles in mid-diastole and end-diastole respectively. Two independent observers described blood flow patterns in the left ventricle using Volume Tracking visualizations.

Results

Volume Tracking was feasible in all eight healthy volunteers and in the patient. Visually, Volume Tracking and particle tracing are complementary methods, showing different aspects of the flow. When validated against particle tracing, on average 90.5% and 87.8% of the particles agreed with the Volume Tracking surface in mid-diastole and end-diastole respectively. Inflow patterns in the left ventricle varied between the subjects, with excellent agreement between observers. The left ventricular inflow pattern in the patient differed from the healthy subjects.

Conclusion

Volume Tracking is a new visualization method for blood flow measured by 4D PC-CMR. Volume Tracking complements and provides incremental information compared to particle tracing that may lead to a better understanding of blood flow and may improve diagnosis and prognosis of cardiovascular diseases.

Improved SNR in phase contrast velocimetry with five-point balanced flow encoding

Phase contrast velocimetry can be utilized to measure complex flow for both quantitative and qualitative assessment of vascular hemodynamics. However, phase contrast requires that a maximum measurable velocity be set that balances noise and phase aliasing. To efficiently reduce noise in phase contrast images, several investigators have proposed extended velocity encoding schemes that use extra encodings to unwrap phase aliasing; however, existing techniques can lead to significant increases in echo and scan time, limiting their clinical benefits. In this work, we have developed a novel five-point velocity encoding scheme that efficiently reduces noise with minimal increases in scan and echo time. Investigations were performed in phantoms, demonstrating a 63% increase in velocity-to-noise ratio compared to standard four-point encoding schemes. Aortic velocity measurements were performed in healthy volunteers, showing similar velocity-to-noise ratio improvements. In those volunteers, it was also demonstrated that, without sacrificing accuracy, low-resolution images can be used for the fifth encoding point, reducing the scan time penalty from 25% down to less than 1%.

Extending the dynamic range of phase contrast magnetic resonance velocity imaging using advanced higher-dimensional phase unwrapping algorithms

Phase contrast magnetic resonance velocity imaging is a powerful technique for quantitative in vivo blood flow measurement. Current practice normally involves restricting the sensitivity of the technique so as to avoid the problem of the measured phase being 'wrapped' onto the range -pi to +pi. However, as a result, dynamic range and signal-to-noise ratio are sacrificed. Alternatively, the true phase values can be estimated by a phase unwrapping process which consists of adding integral multiples of 2pi to the measured wrapped phase values. In the presence of noise and data undersampling, the phase unwrapping problem becomes non-trivial. In this paper, we investigate the performance of three different phase unwrapping algorithms when applied to three-dimensional (two spatial axes and one time axis) phase contrast datasets. A simple one-dimensional temporal unwrapping algorithm, a more complex and robust three-dimensional unwrapping algorithm and a novel velocity encoding unwrapping algorithm which involves unwrapping along a fourth dimension (the 'velocity encoding' direction) are discussed, and results from the three are presented and compared. It is shown that compared to the traditional approach, both dynamic range and signal-to-noise ratio can be increased by a factor of up to five times, which demonstrates considerable promise for a possible eventual clinical implementation. The results are also of direct relevance to users of any other technique delivering time-varying two-dimensional phase images, such as dynamic speckle interferometry and synthetic aperture radar.

Temporal DeltaB0 and relaxation in the rat heart

Field maps of the induced main magnetic field offset (DeltaB(0)) were measured in the rat heart at various points in the cardiac cycle for the purpose of identifying their effects on relaxation measurements. The mean DeltaB(0) of the left ventricle averaged across rats was found to be 0.11 +/- 0.35 ppm and 0.19 +/- 0.39 ppm at the onset of systole and diastole, respectively. The root mean square (RMS) variation in resonant frequency of the left ventricle averaged across rats was found to be 0.09 +/- 0.05 ppm and 0.06 +/- 0.04 ppm during systole and diastole, respectively. Temporal variations in DeltaB(0) could substantially affect quantitative MRI measurements. To assess this, transverse relaxation rates (R(2) and R(2)(*)) were measured at different points in the cardiac cycle, and the effects of DeltaB(0) were estimated using measured field map data. For a given region of the left ventricle, DeltaB(0) induced a mean error across rats of < or =3.9% for R(2) and < or =9.6% for R(2)(*). For R(2)(*) measurements, the static component of the field inhomogeneity was found to be responsible for most of the error induced.

Time-Resolved 3-Dimensional Velocity Mapping in the Thoracic Aorta

OBJECTIVE

An analysis of thoracic aortic blood flow in normal subjects and patients with aortic pathologic findings is presented. Various visualization tools were used to analyze blood flow patterns within a single 3-component velocity volumetric acquisition of the entire thoracic aorta

METHODS

Time-resolved, 3-dimensional phase-contrast magnetic resonance imaging (3D CINE PC MRI) was employed to obtain complete spatial and temporal coverage of the entire thoracic aorta combined with spatially registered 3-directional pulsatile blood flow velocities. Three-dimensional visualization tools, including time-resolved velocity vector fields reformatted to arbitrary 2-dimensional cut planes, 3D streamlines, and time-resolved 3D particle traces, were applied in a study with 10 normal volunteers. Results from 4 patient examinations with similar scan prescriptions to those of the volunteer scans are presented to illustrate flow features associated with common pathologic findings in the thoracic aorta.

RESULTS

Previously reported blood flow patterns in the thoracic aorta, including right-handed helical outflow, late systolic retrograde flow, and accelerated passage through the aortic valve plane, were visualized in all volunteers. The effects of thoracic aortic disease on spatial and temporal blood flow patterns are illustrated in clinical cases, including ascending aortic aneurysms, aortic regurgitation, and aortic dissection.

CONCLUSION

Time-resolved 3D velocity mapping was successfully applied in a study of 10 healthy volunteers and 4 patients with documented aortic pathologic findings and has proven to be a reliable tool for analysis and visualization of normal characteristic as well as pathologic flow features within the entire thoracic aorta.

Understanding phase maps in MRI: a new cutline phase unwrapping method

This paper describes phase maps. A review of the phase unwrapping problem is given. Different structures, in particular fringelines, cutlines, and poles, contained within a phase map are described and their origin and behavior investigated. The problem of phase unwrapping can then be addressed with a better understanding of the source of poles or inconsistencies. This understanding, along with some assumptions about what is being encoded in the phase of a magnetic resonance image, are used to derive a new method for phase unwrapping which relies only on the phase map. The method detects cutlines and distinguishes between noise-induced poles and signal undersampling poles based on the length of the fringelines. The method was shown to be robust to noise and successful in unwrapping challenging clinical cases.

Assessing the accuracy of two-dimensional phase-contrast MRI measurements of complex unsteady flows

Two-dimensional phase-contrast MRI measurements of complex unsteady flows have been assessed for accuracy, together with procedures used to improve the precision of the measurements. Velocity measurements of single harmonic sinusoidal flow in a rigid bypass graft model with a fully three-dimensional geometry were compared to an accurate numerical solution of the Navier-Stokes equations for the same flow. Axial velocity profiles from the MRI were compared with the computational data, and instantaneous root mean square (rms) differences were calculated. Despite the complexity of the flow, with the aid of phase angle dynamic range extension, a spatially and temporally averaged rms error of between 7.8% and 11.5%, with respect to the spatially and temporally averaged velocity, was achieved. Spin saturation primarily and phase dispersion secondarily in complex transient recirculation zones were found to be significant contributors to overall error. Cross flow effects were also investigated but were of lesser significance. The result confirms the suitability of the technique for measuring complex unsteady flows.

Improved estimation of velocity and flow rate using regularized three-point phase-contrast velocimetry

We improved the three-point phase-contrast method by regularization of MR velocity data after acquisition of a low velocity-to-noise ratio (VNR) velocity image and a high VNR aliased velocity image. The phase unwrapping algorithm is based on the assumed correlation of the velocity of adjacent flow voxels on the low VNR and the unaliased high VNR images. We used Fourier encoding with eight velocity-encoding gradient steps to obtain reference velocity images of the aorta from five subjects (274 images) and compared them with the phase-contrast and three-point phase-contrast images with and without regularization. The VNR of the regularized velocity image was improved by 9.1 dB and the VNR of the three-point phase-contrast velocity image was improved by 0.7 dB with respect to the low first moment velocity image. Corresponding improvements of 9 dB and 3.7 dB were obtained for the estimations of instantaneous flow rate. Magn Reson Med 44:122-128, 2000.

Spatial regularization of flow patterns in magnetic resonance velocity mapping

A technique dedicated to spatial regularization of magnetic resonance (MR) velocity data has been implemented to improve flow image quality. It is assumed that neighboring flow-velocity pixels are partially correlated, although large-velocity discontinuities remain possible. Increasing MR signal magnitude due to the in-flow effect also is used to enhance further reliability of the estimated velocity. By using an eight-step Fourier-encoding approach, 162 "reference" velocity images acquired in the ascending aorta from six healthy volunteers were compared with "raw" and "regularized" images that were computed from only two gradient steps. The mean square error decreased from 0.12 m(2) x s(-2) to 0.06 m(2) x s(-2) (P < 10-9) for velocity pixel values and from 1929 ml(2) x s(-2) to 1336 ml(2) x s(-2) (P < 0.01) for instantaneous flow rates. The regularization of two-step data sets provides the same velocity image quality as that found after using three-step data sets without regularization. The method can be applied to phase-velocity data sets of any MR technique to reduce velocity noise. J. Magn. Reson. Imaging 1999;10:851-860.

Particle trace visualization of intracardiac flow using time-resolved 3D phase contrast MRI

The flow patterns in the human heart are complex and difficult to visualize using conventional two-dimensional (2D) modalities, whether they depict a single velocity component (Doppler echocardiography) or all three components in a few slices (2D phase contrast MRI). To avoid these shortcomings, a temporally resolved 3D phase contrast technique was used to derive data describing the intracardiac velocity fields in normal volunteers. The MRI data were corrected for phase shifts caused by eddy currents and concomitant gradient fields, with improvement in the accuracy of subsequent flow visualizations. Pathlines describing the blood pathways through the heart were generated from the temporally resolved velocity data, starting from user-specified locations and time frames. Flow trajectories were displayed as 3D particle traces, with simultaneous demonstration of morphologic 2D slices. This type of visualization is intuitive and interactive and may extend our understanding of dynamic and previously unrecognized patterns of intracardiac flow.