Michael Markl

Purpose: Hemodynamic monitoring is essential for patients with right‐sided congenital heart disease (CHD). Respiration may have an increased impact on pulmonary flow in these patients that cannot be assessed by standard tools including 4D flow MRI. This study uses 5D flow MRI to assess respiratory‐cycle variations in flow energetics in patients with CHD. Methods: 5D flow was acquired with four respiratory states in 14 Fontan patients (21 ± 8 years, 8 female), 10 intracardiac shunt patients (19 ± 13 years, 8 female), and 9 controls (26 ± 6 years, 1 female). Blood kinetic energy (KE _mean ), viscous energy loss (EL _total ), and EL fraction (EL _total /KE _mean ) as a measure of flow inefficiency were calculated in inferior and superior caval veins (IVC, SVC), pulmonary arteries (PA), and aorta. Correlations were assessed with clinical markers of altered cardiac flow function. Results: 5D flow was acquired with acceleration factor R = 37–93 varying between respiratory states. Fontan and shunt patients demonstrated significant respiratory‐driven changes in pulmonary flow energetics compared to controls. Fontan IVC and PA KE _mean were increased during inspiration (+43%, +37%, p < 0.001) and decreased during expiration (−40%, −35%, p < 0.001), resulting in increased expiratory EL fraction (+34%, +30%, p < 0.05). Shunt patients showed a similar effect in IVC KE _mean (+28%, −25%, p < 0.05). Decreased expiratory IVC KE _mean was associated with increased Fontan left–right PA flow differential (ρ = −0.68, p < 0.05) and increased shunt Qp/Qs (ρ = −0.70, p < 0.05). Conclusions: The findings of this study show that CHD flow energetics and efficiency are modulated by respiration. Respiratory‐resolved imaging is needed to identify these dynamics and their relationships to overall cardiac function.

Background: Accurate anatomical labeling of intracranial arteries is critical for cerebrovascular diagnosis and hemodynamic analysis, but remains time-consuming and prone to inter-operator variability. While deep learning provides an automated solution, its clinical adoption is limited by the lack of confidence measures. Incorporating uncertainty quantification into automated labeling could enhance interpretability by identifying ambiguous or abnormal regions and support clinical trust, yet this aspect remains underexplored. Methods: To address this gap, we introduce an uncertainty-aware deep learning framework for automated artery labeling from 3D Time-of-Flight Magnetic Resonance Angiography (3D ToF-MRA) segmentations ( n = 35). Three convolutional neural network architectures were evaluated: (1) UNet with residual encoder blocks, (2) CS-Net, an attention-augmented UNet with spatial attention, and (3) nnUNet, a self-configuring framework that adapts architecture and training to dataset characteristics. Confidence was modeled via test-time augmentation (TTA) combined with a novel coordinate-guided strategy to reduce interpolation errors during inference. Generalizability was assessed by evaluating a subset of the public TubeTK ToF-MRA dataset ( n = 20). Results: Voxelwise uncertainty maps highlighted anatomical ambiguities, pathological variations, and inconsistencies in manual references, providing intuitive confidence indicators. nnUNet achieved the highest performance (average Dice score 0.93; clDice 0.94; average surface distance 0.35 mm; 95th percentile of Hausdorff distance 4.51 mm), demonstrating robustness in complex vascular regions. On the TubeTK dataset, nnUNet maintained robust generalization (average Dice score 0.87; clDice 0.87; average surface distance 0.42 mm; 95th percentile of Hausdorff distance 5.85 mm). Validation against co-registered 4D flow MRI showed close agreement between flow velocities derived from automated and manual labels, with no significant differences. Conclusion: The proposed framework delivers a scalable, accurate, and uncertainty-aware solution for intracranial artery labeling. By integrating uncertainty quantification, it offers a transparent and clinically trustworthy tool to facilitate cerebrovascular imaging workflows and support subsequent hemodynamic analyses.

Chronic thromboembolic pulmonary hypertension (CTEPH) can lead to right ventricular (RV) ischemia and dysfunction due to chronic pulmonary artery obstruction and increased afterload. While cardiovascular MRI (CMR) enables noninvasive assessment of myocardial perfusion, its role in CTEPH remains unclear. The authors report adenosine stress perfusion CMR findings from two patients with CTEPH before and after pulmonary thromboendarterectomy (PTE). Both showed reduced biventricular perfusion before PTE; one demonstrated post-PTE improvement. Perfusion findings aligned with invasive hemodynamics, suggesting that CMR-derived myocardial perfusion reserve may serve as a valuable tool for assessing treatment response and RV pathophysiologic characteristics in CTEPH. Keywords: Cardiac, Pulmonary Arteries, Chronic Thromboembolic Pulmonary Hypertension, Pulmonary Thromboendarterectomy, Quantitative Perfusion Cardiovascular MRI, Myocardial Blood Flow, Myocardial Perfusion Reserve Supplemental material is available for this article. © RSNA, 2025

(1) Objective: The objective was to test the hypothesis that cine MRI-derived radiomic features can detect functional tricuspid regurgitation (FTR) in the context of pulmonary hypertension (PH). (2) Materials and methods: In total, 53 PH patients were retrospectively enrolled. Thirty-three patients had echocardiography-defined mild-to-severe FTR, while the other twenty patients had no or trivial regurgitation. For all participants, 93 radiomic features were extracted from four-chamber cine MRI using a fixed-size region of interest (ROI) located in the right atrium (RA), 0.5–1 cm above the tricuspid valve. The levels of radiomic features were averaged over the ventricular systole and compared between patients with and without FTR using t tests. In patients with FTR, radiomic features were related to hemodynamic parameters in the right heart using the Pearson correlation coefficient (r). (3) Results: There were no significant differences in demographic information, right heart catheterization (RHC) results, and most cine MRI-derived cardiac function indices between the two subject groups. Eight of ninety-three radiomic features were significantly different between PH patients with and without FTR. Radiomic features can be used to discriminate two subject groups (AUC = 0.77). In patients with FTR, multiple radiomic features are related to the pressure in the RA, right ventricle (RV), and pressure difference between RA and RV (r: 0.4 to 0.55), p values < 0.05. (4) Conclusion: Cine MRI-derived radiomic features of the cardiac blood pool differ between PH patients with and without FTR. Cine MRI shows promise as a method for assessing FTR in the context of complex cardiovascular diseases (CVDs).

This review highlights selected applications and future directions of four-dimensional flow MRI in the thoracic aorta.

Four-dimensional (4D) flow MRI has shown promise for the assessment of aortic hemodynamics. However, data analysis traditionally requires manual and time-consuming human input at several stages. This limits reproducibility and affects analysis workflows, such that large-cohort 4D flow studies are lacking. Here, a fully automated artificial intelligence (AI) 4D flow analysis pipeline was developed and evaluated in a cohort of over 350 subjects. The 4D flow MRI analysis pipeline integrated a series of previously developed and validated deep learning networks, which replaced traditionally manual processing tasks (background-phase correction, noise masking, velocity anti-aliasing, aorta 3D segmentation). Hemodynamic parameters (global aortic pulse wave velocity (PWV), peak velocity, flow energetics) were automatically quantified. The pipeline was evaluated in a heterogeneous single-center cohort of 379 subjects (age = 43.5 ± 18.6 years, 118 female) who underwent 4D flow MRI of the thoracic aorta (n = 147 healthy controls, n = 147 patients with a bicuspid aortic valve [BAV], n = 10 with mechanical valve prostheses, n = 75 pediatric patients with hereditary aortic disease). Pipeline performance with BAV and control data was evaluated by comparing to manual analysis performed by two human observers. A fully automated 4D flow pipeline analysis was successfully performed in 365 of 379 patients (96%). Pipeline-based quantification of aortic hemodynamics was closely correlated with manual analysis results (peak velocity: r = 1.00, p < 0.001; PWV: r = 0.99, p < 0.001; flow energetics: r = 0.99, p < 0.001; overall r ≥ 0.99, p < 0.001). Bland–Altman analysis showed close agreement for all hemodynamic parameters (bias 1–3%, limits of agreement 6–22%). Notably, limits of agreement between different human observers’ quantifications were moderate (4–20%). In addition, the pipeline 4D flow analysis closely reproduced hemodynamic differences between age-matched adult BAV patients and controls (median peak velocity: 1.74 m/s [automated] or 1.76 m/s [manual] BAV vs. 1.31 [auto.] vs. 1.29 [manu.] controls, p < 0.005; PWV: 6.4–6.6 m/s all groups, any processing [no significant differences]; kinetic energy: 4.9 μJ [auto.] or 5.0 μJ [manu.] BAV vs. 3.1 μJ [both] control, p < 0.005). This study presents a framework for the complete automation of quantitative 4D flow MRI data processing with a failure rate of less than 5%, offering improved measurement reliability in quantitative 4D flow MRI. Future studies are warranted to reduced failure rates and evaluate pipeline performance across multiple centers.

A fluid physics–informed cycle generative adversarial network was successfully used to derive accurate three-dimensional aorta hemodynamics in less than 1 second using only aorta anatomic features as input data.

Stenosis of major intracranial arteries is a significant cause of stroke, with assessment of trans-stenotic pressure drops being a key marker of functional stenosis severity. Non-invasive methods for quantifying intracranial pressure changes are hence crucial; however, the narrow and tortuous cerebrovascular network poses challenges to traditional assessment methods such as transcranial Doppler. This study investigates the use of novel deep learning-enhanced super-resolution (SR) four-dimensional (4D) flow magnetic resonance imaging (MRI) in combination with a physics-informed virtual work–energy relative pressure technique to quantify pressure drops across stenotic intracranial arteries. Performance was validated in intracranial-mimicking in vitro experiments using pulsatile flow before being transferred into an in vivo cohort of patients with intracranial atherosclerotic disease. Conversion into sub-millimetre SR imaging significantly improved the accuracy of regional relative pressure estimations in the pulsing brain arteries, mitigating biases observed at >1 mm resolution imaging, and agreeing strongly with reference catheter-based invasive measurements across both moderate and severe stenoses. The in vivo analysis also revealed a significant increase in pressure drops when converting into sub-millimetre SR data, underlining the importance of apparent image resolution in a clinical setting. The results highlight the potential of SR 4D flow MRI for non-invasive quantification of cerebrovascular pressure changes in pulsing intracranial arteries across stenotic vessel segments.

The released CMRxRecon2024 dataset is currently the largest and most protocol-diverse publicly available k-space dataset including multimodality and multiview cardiac MRI data from 330 healthy volunteers, and each one covers standardized and commonly used clinical protocols.