4DFlowNet: Super-Resolution 4D Flow MRI Using Deep Learning and Computational Fluid Dynamics

Ferdian, Edward; Suinesiaputra, Avan; Dubowitz, David J.; Zhao, Debbie; Wang, Alan; Cowan, Brett R.; Young, Alistair A.

doi:10.3389/fphy.2020.00138

Cited by 95 publications

(75 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The proposed DL framework required complicated methodology to generate training data. A simulated training data approach (ie, CFD generated flow profiles) as used by Ferdian et al 22 may serve as a simpler alternative for generating DL training data. Despite our best efforts, a large portion of the total reconstruction time for CS and DL was due to pre‐processing time [DL = 40.9 s (88.0% of total recon time), CS = 62.6 s (30.0 % of total recon time)].…”

Section: Discussionmentioning

confidence: 99%

“…Vishnevskiy et al 21 showed that an unrolled network incorporating a physics‐based model into the DL architecture reduces reconstruction time 30‐fold compared to CS for 12.4‐ to 13.8‐fold accelerated 4D ECG‐gated segmented PC MRI with 25 cardiac phases. Ferdian et al 22 showed that 4DFlowNet network trained using synthetic 4D flow MR generated from computational fluid dynamic (CFD) solutions could be used to increase spatial resolution. Nath et al 23 showed that a U‐net could filter aliasing artifact from accelerated (2.5‐fold ≤ acceleration rate ≤ 5‐fold) 2D PC MRI, while outperforming CS.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Highly accelerated free‐breathing real‐time phase contrast cardiovascular MRI via complex‐difference deep learning

Haji‐Valizadeh

Guo

Küçükseymen

et al. 2021

Magnetic Resonance in Med

View full text Add to dashboard Cite

Purpose: To develop and evaluate a real-time phase contrast (PC) MRI protocol via complex-difference deep learning (DL) framework. Methods: DL used two 3D U-nets to separately filter aliasing artifact from radial real-time velocity-compensated and complex-difference images. U-nets were trained with synthetic real-time PC generated from electrocardiograph (ECG) -gated, breathhold, segmented PC (ECG-gated segmented PC) acquired at the ascending aorta of 510 patients. In 21 patients, free-breathing, ungated real-time (acceleration rate = 28.8) and ECG-gated segmented (acceleration rate = 2) PC were prospectively acquired at the ascending aorta. Hemodynamic parameters (cardiac output [CO], stroke volume [SV], and mean velocity at peak systole [peak mean velocity]) were measured for ECG-gated segmented and DL-filtered synthetic real-time PC and compared using Bland-Altman and linear regression analyses. Additionally, hemodynamic parameters were quantified from DL-filtered, compressed-sensing (CS) -reconstructed, and gridding reconstructed prospective real-time PC and compared to ECG-gated segmented PC. Results: Synthetic real-time PC with DL showed strong correlation (R > 0.98) and good agreement with ECG-gated segmented PC for quantified hemodynamic parameters (mean-difference: CO = −0.3 L/min, SV = −4.3 mL, peak mean velocity = −2.3 cm/s). On average, DL required 0.39 s/frame to filter prospective real-time PC, which was 4.6-fold faster than CS. Compared to CS, DL showed superior correlation, tighter limits of agreement (LOAs), better bias for peak mean velocity, and worse bias for CO and SV. Compared to gridding, DL showed similar correlation, tighter LOAs for CO and SV, similar bias for CO, and worse bias for SV and peak mean velocity. Conclusion:The complex-difference DL framework accelerated real-time PC-MRI by nearly 28-fold, enabling rapid free-running real-time assessment of flow hemodynamics.How to cite this article: Haji-Valizadeh H, Guo R, Kucukseymen S, et al. Highly accelerated free-breathing real-time phase contrast cardiovascular MRI via complex-difference deep learning.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Highly accelerated free‐breathing real‐time phase contrast cardiovascular MRI via complex‐difference deep learning

Haji‐Valizadeh

Guo

Küçükseymen

et al. 2021

Magnetic Resonance in Med

View full text Add to dashboard Cite

show abstract

“…As a result, reliable imaging of the LAA flow field is extremely challenging, especially in the proximity of the vessel wall, making it nearly impossible to obtain accurate values of derived hemodynamic indices such as the wall shear stress or the ECAP (Petersson et al, 2012 ). In this regard, attempts have already been made to tackle said limitations such as the development of Dual-V enc acquisition sequences (Callahan et al, 2019 ) or leveraging CFD simulations to obtain 4D flow super-resolution (Ferdian et al, 2020 ).…”

Section: Discussionmentioning

confidence: 99%

Deep Learning Framework for Real-Time Estimation of in-silico Thrombotic Risk Indices in the Left Atrial Appendage

et al. 2021

View full text Add to dashboard Cite

Patient-specific computational fluid dynamics (CFD) simulations can provide invaluable insight into the interaction of left atrial appendage (LAA) morphology, hemodynamics, and the formation of thrombi in atrial fibrillation (AF) patients. Nonetheless, CFD solvers are notoriously time-consuming and computationally demanding, which has sparked an ever-growing body of literature aiming to develop surrogate models of fluid simulations based on neural networks. The present study aims at developing a deep learning (DL) framework capable of predicting the endothelial cell activation potential (ECAP), an in-silico index linked to the risk of thrombosis, typically derived from CFD simulations, solely from the patient-specific LAA morphology. To this end, a set of popular DL approaches were evaluated, including fully connected networks (FCN), convolutional neural networks (CNN), and geometric deep learning. While the latter directly operated over non-Euclidean domains, the FCN and CNN approaches required previous registration or 2D mapping of the input LAA mesh. First, the superior performance of the graph-based DL model was demonstrated in a dataset consisting of 256 synthetic and real LAA, where CFD simulations with simplified boundary conditions were run. Subsequently, the adaptability of the geometric DL model was further proven in a more realistic dataset of 114 cases, which included the complete patient-specific LA and CFD simulations with more complex boundary conditions. The resulting DL framework successfully predicted the overall distribution of the ECAP in both datasets, based solely on anatomical features, while reducing computational times by orders of magnitude compared to conventional CFD solvers.

show abstract

“…In general, the family of Kalman filter data assimilation models provides the means to improve the accuracy of computational models by leveraging even imperfect experimental data. There are a growing number of studies in the literature on merging CFD and 4D flow MRI data using data assimilation [70,[77][78][79][80][81]. PIV is another popular approach in haemodynamics quantification.…”

Section: Opportunities and Challengesmentioning

confidence: 99%

Data-driven cardiovascular flow modelling: examples and opportunities

Arzani

Dawson

2021

J. R. Soc. Interface.

View full text Add to dashboard Cite

High-fidelity blood flow modelling is crucial for enhancing our understanding of cardiovascular disease. Despite significant advances in computational and experimental characterization of blood flow, the knowledge that we can acquire from such investigations remains limited by the presence of uncertainty in parameters, low resolution, and measurement noise. Additionally, extracting useful information from these datasets is challenging. Data-driven modelling techniques have the potential to overcome these challenges and transform cardiovascular flow modelling. Here, we review several data-driven modelling techniques, highlight the common ideas and principles that emerge across numerous such techniques, and provide illustrative examples of how they could be used in the context of cardiovascular fluid mechanics. In particular, we discuss principal component analysis (PCA), robust PCA, compressed sensing, the Kalman filter for data assimilation, low-rank data recovery, and several additional methods for reduced-order modelling of cardiovascular flows, including the dynamic mode decomposition and the sparse identification of nonlinear dynamics. All techniques are presented in the context of cardiovascular flows with simple examples. These data-driven modelling techniques have the potential to transform computational and experimental cardiovascular research, and we discuss challenges and opportunities in applying these techniques in the field, looking ultimately towards data-driven patient-specific blood flow modelling.

show abstract

4DFlowNet: Super-Resolution 4D Flow MRI Using Deep Learning and Computational Fluid Dynamics

Cited by 95 publications

References 25 publications

Highly accelerated free‐breathing real‐time phase contrast cardiovascular MRI via complex‐difference deep learning

Highly accelerated free‐breathing real‐time phase contrast cardiovascular MRI via complex‐difference deep learning

Deep Learning Framework for Real-Time Estimation of in-silico Thrombotic Risk Indices in the Left Atrial Appendage

Data-driven cardiovascular flow modelling: examples and opportunities

Contact Info

Product

Resources

About