Search

Pulmonary artery (PA) flow analysis is crucial for understanding the progression of pulmonary hypertension (PH). We hypothesized that PA flow characteristics vary according to PH etiology. In this study, we used 4D flow cardiovascular magnetic resonance imaging (CMR) to compare PA flow velocity and wall shear stress (WSS) between patients with pulmonary arterial hypertension (PAH) and those with heart failure with preserved ejection fraction and pulmonary hypertension (PH-HFpEF). We enrolled 13 PAH and 15 PH-HFpEF patients. All participants underwent echocardiography, 4D flow CMR, and right heart catheterization. We compared right ventricular outflow tract (RVOT) flow and main pulmonary artery (MPA) hemodynamics, including peak velocity and mean and maximum WSS, between groups. PH-HFpEF patients were older and more likely to have hypertension. PAH patients had higher mean PA pressure (47.8 ± 8.8 vs. 32.9 ± 6.9 mmHg, p  < 0.001) and pulmonary vascular resistance (PVR) (8.6 ± 4.6 vs. 2.6 ± 2.2 wood unit, p  < 0.001). RVOT systolic notching was more common in PAH patients (8 of 13 vs. 0 of 15), and they had shorter RVOT acceleration time (85.5 ± 20.9 vs. 135.0 ± 21.7 ms, p  < 0.001). PAH patients had lower MPA Vmax (0.8 ± 0.2 vs. 1.1 ± 0.4 m/s, p  = 0.032), mean WSS (0.29 ± 0.09 vs. 0.36 ± 0.06 Pa, p  = 0.035), and maximal WSS (0.99 ± 0.18 vs. 1.21 ± 0.19 Pa, p  = 0.011). Anterior MPA analysis confirmed lower WSS in PAH patients. PVR was negatively correlated with MPA mean WSS ( r  = -0.630, p  = 0.002). PAH patients had lower MPA Vmax and lower mean MPA WSS in 4D flow CMR compared to PH-HFpEF patients. These distinct PA flow characteristics suggest that the flow hemodynamics of the PA remodeling process differ depending on the underlying etiology of PH., Competing Interests: The authors declare no conflicts of interest., (© 2024 The Author(s). Pulmonary Circulation published by John Wiley & Sons Ltd on behalf of Pulmonary Vascular Research Institute.)

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., (© 2022 The Authors. Magnetic Resonance in Medicine published by Wiley Periodicals LLC on behalf of International Society for Magnetic Resonance in Medicine.)

Tissue phase mapping (TPM) is an MRI technique for quantification of regional biventricular myocardial velocities. Despite its potential, clinical use is limited due to the requisite labor-intensive manual segmentation of cardiac contours for all time frames. The purpose of this study was to develop a deep learning (DL) network for automated segmentation of TPM images, without significant loss in segmentation and myocardial velocity quantification accuracy compared with manual segmentation. We implemented a multi-channel 3D (three dimensional; 2D + time) dense U-Net that trained on magnitude and phase images and combined cross-entropy, Dice, and Hausdorff distance loss terms to improve the segmentation accuracy and suppress unnatural boundaries. The dense U-Net was trained and tested with 150 multi-slice, multi-phase TPM scans (114 scans for training, 36 for testing) from 99 heart transplant patients (44 females, 1-4 scans/patient), where the magnitude and velocity-encoded (V x , V y , V z ) images were used as input and the corresponding manual segmentation masks were used as reference. The accuracy of DL segmentation was evaluated using quantitative metrics (Dice scores, Hausdorff distance) and linear regression and Bland-Altman analyses on the resulting peak radial and longitudinal velocities (V r and V z ). The mean segmentation time was about 2 h per patient for manual and 1.9 ± 0.3 s for DL. Our network produced good accuracy (median Dice = 0.85 for left ventricle (LV), 0.64 for right ventricle (RV), Hausdorff distance = 3.17 pixels) compared with manual segmentation. Peak V r and V z measured from manual and DL segmentations were strongly correlated (R ≥ 0.88) and in good agreement with manual analysis (mean difference and limits of agreement for V z and V r were -0.05 ± 0.98 cm/s and -0.06 ± 1.18 cm/s for LV, and -0.21 ± 2.33 cm/s and 0.46 ± 4.00 cm/s for RV, respectively). The proposed multi-channel 3D dense U-Net was capable of reducing the segmentation time by 3,600-fold, without significant loss in accuracy in tissue velocity measurements., (© 2021 John Wiley & Sons, Ltd.)

Purpose: To systematically assess the feasibility and performance of a highly accelerated compressed sensing (CS) 4D flow MRI framework at three different acceleration factors (R) for the quantification of aortic flow dynamics and wall shear stress (WSS) in patients with aortic disease., Methods: Twenty patients with aortic disease (58 ± 15 y old; 19 M) underwent four 4D flow scans: one conventional (GRAPPA, R = 2) and three CS 4D flows with R = 5.7, 7.7, and 10.2. All scans were acquired with otherwise equivalent imaging parameters on a 1.5T scanner. Peak-systolic velocity (V max ), peak flow (Q max ), and net flow (Q net ) were quantified at the ascending aorta (AAo), arch, and descending aorta (DAo). WSS was calculated at six regions within the AAo and arch., Results: Mean scan times for the conventional and CS 4D flows with R = 5.7, 7.7, and 10.2 were 9:58 ± 2:58 min, 3:40 ± 1:19 min, 2:50 ± 0:56 min, and 2:05 ± 0:42 min, respectively. V max , Q max , and Q net were significantly underestimated by all CS protocols (underestimation ≤ -7%, -9%, and -10% by CS, R = 5.7, 7.7, and 10.2, respectively). WSS measurements showed the highest underestimation by all CS protocols (underestimation ≤ -9%, -12%, and -14% by CS, R = 5.7, 7.7, and 10.2)., Conclusions: Highly accelerated aortic CS 4D flow at R = 5.7, 7.7, and 10.2 showed moderate agreement with the conventional 4D flow, despite systematically underestimating various hemodynamic parameters. The shortened scan time may enable the clinical translation of CS 4D flow, although potential hemodynamic underestimation should be considered when interpreting the results., (© 2020 International Society for Magnetic Resonance in Medicine.)

Purpose: To generate fully automated and fast 4D-flow MRI-based 3D segmentations of the aorta using deep learning for reproducible quantification of aortic flow, peak velocity, and dimensions., Methods: A total of 1018 subjects with aortic 4D-flow MRI (528 with bicuspid aortic valve, 376 with tricuspid aortic valve and aortic dilation, 114 healthy controls) comprised the data set. A convolutional neural network was trained to generate 3D aortic segmentations from 4D-flow data. Manual segmentations served as the ground truth (N = 499 training, N = 101 validation, N = 418 testing). Dice scores, Hausdorff distance, and average symmetrical surface distance were calculated to assess performance. Aortic flow, peak velocity, and lumen dimensions were quantified at the ascending, arch, and descending aorta and compared using Bland-Altman analysis. Interobserver variability of manual analysis was assessed on a subset of 40., Results: Convolutional neural network segmentation required 0.438 ± 0.355 seconds versus 630 ± 254 seconds for manual analysis and demonstrated excellent performance with a median Dice score of 0.951 (0.930-0.966), Hausdorff distance of 2.80 (2.13-4.35), and average symmetrical surface distance of 0.176 (0.119-0.290). Excellent agreement was found for flow, peak velocity, and dimensions with low bias and limits of agreement less than 10% difference versus manual analysis. For aortic volume, limits of agreement were moderate within 16.3%. Interobserver variability (median Dice score: 0.950; Hausdorff distance: 2.45; and average symmetrical surface distance: 0.145) and convolutional neural network-based analysis (median Dice score: 0.953-0.959; Hausdorff distance: 2.24-2.91; and average symmetrical surface distance: 0.145-1.98 to observers) demonstrated similar reproducibility., Conclusions: Deep learning enabled fast and automated 3D aortic segmentation from 4D-flow MRI, demonstrating its potential for efficient clinical workflows. Future studies should investigate its utility for other vasculature and multivendor applications., (© 2020 International Society for Magnetic Resonance in Medicine.)