scholarly journals Cerebrovascular super-resolution 4D Flow MRI – using deep learning to non-invasively quantify velocity, flow, and relative pressure

2021 ◽  
Author(s):  
E. Ferdian ◽  
D. Marlevi ◽  
J. Schollenberger ◽  
M. Aristova ◽  
E.R. Edelman ◽  
...  
Keyword(s):  
Deep Learning ◽  
Relative Error ◽  
Super Resolution ◽  
Patient Specific ◽  
Relative Pressure ◽  
4D Flow Mri ◽  
4D Flow ◽  
Velocity Flow ◽  
Flow Mri ◽  
Cerebrovascular Hemodynamics

ABSTRACTThe development of cerebrovascular disease is tightly coupled to changes in cerebrovascular hemodynamics, with altered flow and relative pressure indicative of the onset, development, and acute manifestation of pathology. Image-based monitoring of cerebrovascular hemodynamics is, however, complicated by the narrow and tortuous vasculature, where accurate output directly depends on sufficient spatial resolution. To address this, we present a method combining dedicated deep learning and state-of-the-art 4D Flow MRI to generate super-resolution full-field images with coupled quantification of relative pressure using a physics-driven image processing approach. The method is trained and validated in a patient-specific in-silico cohort, showing good accuracy in estimating velocity (relative error: 12.0 ± 0.1%, mean absolute error (MAE): 0.07 ± 0.06 m/s at peak velocity), flow (relative error: 6.6 ± 4.7%, root mean square error (RMSE): 0.5 ± 0.1 mL/s at peak flow), and with maintained recovery of relative pressure through the circle of Willis (relative error: 11.0 ± 7.3%, RMSE: 0.3 ± 0.2 mmHg). Furthermore, the method is applied to an in-vivo volunteer cohort, effectively generating data at <0.5mm resolution and showing potential in reducing low-resolution bias in relative pressure estimation. Our approach presents a promising method to non-invasively quantify cerebrovascular hemodynamics, applicable to dedicated clinical cohorts in the future.

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

2020 ◽  
Vol 8 ◽  
Author(s):  
Edward Ferdian ◽  
Avan Suinesiaputra ◽  
David J. Dubowitz ◽  
Debbie Zhao ◽  
Alan Wang ◽  
...  
Keyword(s):  
Fluid Dynamics ◽  
Computational Fluid Dynamics ◽  
Deep Learning ◽  
Super Resolution ◽  
4D Flow Mri ◽  
4D Flow ◽  
Flow Mri

scholarly journals Enhancement of cerebrovascular 4D flow MRI velocity fields using machine learning and computational fluid dynamics simulation data

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
David R. Rutkowski ◽  
Alejandro Roldán-Alzate ◽  
Kevin M. Johnson
Keyword(s):  
Fluid Dynamics ◽  
Computational Fluid Dynamics ◽  
Velocity Fields ◽  
Dynamics Simulation ◽  
Patient Specific ◽  
4D Flow Mri ◽  
4D Flow ◽  
Flow Phase ◽  
Flow Mri ◽  
Trained Network

AbstractBlood flow metrics obtained with four-dimensional (4D) flow phase contrast (PC) magnetic resonance imaging (MRI) can be of great value in clinical and experimental cerebrovascular analysis. However, limitations in both quantitative and qualitative analyses can result from errors inherent to PC MRI. One method that excels in creating low-error, physics-based, velocity fields is computational fluid dynamics (CFD). Augmentation of cerebral 4D flow MRI data with CFD-informed neural networks may provide a method to produce highly accurate physiological flow fields. In this preliminary study, the potential utility of such a method was demonstrated by using high resolution patient-specific CFD data to train a convolutional neural network, and then using the trained network to enhance MRI-derived velocity fields in cerebral blood vessel data sets. Through testing on simulated images, phantom data, and cerebrovascular 4D flow data from 20 patients, the trained network successfully de-noised flow images, decreased velocity error, and enhanced near-vessel-wall velocity quantification and visualization. Such image enhancement can improve experimental and clinical qualitative and quantitative cerebrovascular PC MRI analysis.

Deep Learning Automated Background Phase Error Correction for Abdominopelvic 4D Flow MRI

Radiology ◽  
2021 ◽  
Author(s):  
Sophie You ◽  
Evan M. Masutani ◽  
Marcus T. Alley ◽  
Shreyas S. Vasanawala ◽  
Pam R. Taub ◽  
...  
Keyword(s):  
Deep Learning ◽  
Error Correction ◽  
Phase Error ◽  
4D Flow Mri ◽  
4D Flow ◽  
Flow Mri ◽  
Background Phase

Abstract 14534: In vitro Assessment of Transvalvular Peak Velocity Measurement for Various Orifice Shapes: Comparison Between 4D Flow MRI vs. Doppler Echocardiography

Circulation ◽  
2020 ◽  
Vol 142 (Suppl_3) ◽  
Author(s):  
Jeesoo Lee ◽  
Nadia El hangouche ◽  
Liliana Ma ◽  
Michael Scott ◽  
Michael Markl ◽  
...  
Keyword(s):  
Temporal Resolution ◽  
Continuous Wave ◽  
Peak Velocity ◽  
Left Ventricular ◽  
Patient Specific ◽  
Maximal Velocity ◽  
4D Flow Mri ◽  
4D Flow ◽  
Flow Phantom ◽  
Flow Mri

Introduction: 4D flow MRI can assess transvalvular velocity, but validation against continuous wave (CW) Doppler echo is limited in high-velocity regurgitation and stenosis situations. We sought to compare 4D flow MRI and echo peak velocity using a pulsatile echo-MRI flow phantom. Materials and Methods: An MRI-compatible flow phantom with restrictive orifice situated was driven by a left ventricular assist device at 50 bpm (figure 1A). Three orifice shapes were tested: circular, elliptical and 3D-printed patient-specific mitral regurgitant orifice model of prolapse with areas of 0.5, 0.41 and 0.35 cm 2 , respectively. CW Doppler was acquired with peak velocity extracted from the profile. Retrospectively-gated 4D flow MRI was performed (spatial resolution = 2 mm isotropic, temporal resolution = 36 ms, encoding velocity = 400 cm/s). Maximal velocity magnitude was extracted volumetrically (figure 1B). An echo-mimicking profile was also obtained with a “virtual” ultrasound beam in the 4D flow data to simulate CW Doppler (figure 1C). Bland-Altman analysis was used to assess the agreement of temporal peak velocities. Results: 4D flow MRI demonstrated a centrally directed jet for the circular and elliptical orifices and an oblique jet for the prolapse orifice (figure 1B). Peak velocities were in excellent agreement between 4D flow MRI vs. echo for the circular (peak: 5.13 vs. 5.08 m/s, bias = 0.06 ± 0.66 m/s, figure 1D) and the elliptical orifice (peak: 4.95 vs. 4.79 m/s, bias = 0.07 ± 0.87 m/s, figure 1E). The prolapse orifice velocity was underestimated somewhat by MRI by ~10% (peak: 4.41 vs. 4.90 m/s, bias=0.26±1.18, figure 1F). Conclusion: 4D flow MRI can quantify high velocities like echo for simple geometries while underestimating for more complex geometry, likely due to partial volume effects. Further investigation is warranted to systematically investigate the effects of 4D flow MRI spatial and temporal resolution as well as the jet angle on velocity quantification accuracy.

Super-resolution and denoising of 4D-Flow MRI using physics-Informed deep neural nets

2020 ◽  
Vol 197 ◽  
pp. 105729
Author(s):  
Mojtaba F. Fathi ◽  
Isaac Perez-Raya ◽  
Ahmadreza Baghaie ◽  
Philipp Berg ◽  
Gabor Janiga ◽  
...  
Keyword(s):  
Super Resolution ◽  
Neural Nets ◽  
4D Flow Mri ◽  
4D Flow ◽  
Flow Mri

Abstract 14689: Regional Aortic Wall Shear Stress Mapping Implicates Hemodynamics in Human Bicuspid Aortopathy

Circulation ◽  
2014 ◽  
Vol 130 (suppl_2) ◽  
Author(s):  
David G Guzzardi ◽  
Pim van Ooij ◽  
Alex J Barker ◽  
Giampaolo Martufi ◽  
Katherine E Olsen ◽  
...  
Keyword(s):  
Shear Stress ◽  
Wall Shear Stress ◽  
Aortic Wall ◽  
Transforming Growth Factor ◽  
Patient Specific ◽  
4D Flow Mri ◽  
4D Flow ◽  
Hemodynamic Assessment ◽  
Flow Mri ◽  
Wall Shear

Introduction: A suspected genetic cause for bicuspid aortic valve (BAV) aortopathy has led to aggressive resection strategies. Using 4D flow MRI, we documented increased regional wall shear stress (WSS) in BAV patients. Local hemodynamics may exacerbate extracellular matrix (ECM) degradation leading to disease progression. If validated, preoperative regional hemodynamic assessment could be used to guide more targeted patient-specific aortic resection. For the first time, we correlated regional WSS with aortic tissue remodeling in BAV patients. Methods & Results: BAV patients (N=11) undergoing ascending aortic resection received preoperative 4D flow MRI with regional WSS differences mapped. Paired aortic wall samples (from same-patient with elevated WSS paired to normal WSS regions) were collected during surgery and compared using histology (pentachrome), biomechanics (biaxial mechanical testing), and ECM regulation (protein expression). Patient mean age: 49±18 years; mean aortic diameter: 4.6±0.7cm (range: 3.6 - 6.3cm); 55% had R+L fusion pattern; 36% had severe aortic stenosis. All patients had heterogeneous WSS patterns with regions of elevated WSS adjacent to those of normal WSS. By histology, regions of increased WSS showed greater medial elastin fragmentation, fibrosis, and cystic medial necrosis compared to adjacent areas of normal WSS. Regions of increased WSS showed increased elastic modulus (fold change±SD: 1.53±0.68; P=0.06, N=5) and collagen stiffness (1.37±0.49; P=0.07, N=5) compared to normal WSS regions suggesting altered distensibility. Multiplex protein analyses of ECM regulatory molecules revealed an increase in transforming growth factor β-1 (1.49±0.71, P=0.02), MMP-1 (1.62±0.84; P=0.01), MMP-2 (1.49±1.00; P=0.06), MMP-3 (1.23±0.36; P=0.02), MMP-7 (1.57±0.75; P=0.02), and TIMP-2 (1.26±0.33; P=0.01) in elevated WSS regions suggesting ECM dysregulation consistent with aortic remodeling. Conclusions: In BAV aorta, regional WSS corresponds with local histologic abnormalities, altered biomechanics, and ECM dysregulation. These novel data strongly implicate local hemodynamics as a mediator of BAV aortopathy. With further validation, 4D flow MRI could be used to guide personalized resection strategies.

scholarly journals Assessment of turbulent blood flow and wall shear stress in aortic coarctation using image-based simulations

2021 ◽  
Vol 20 (1) ◽  
Author(s):  
Romana Perinajová ◽  
Joe F. Juffermans ◽  
Jonhatan Lorenzo Mercado ◽  
Jean-Paul Aben ◽  
Leon Ledoux ◽  
...  
Keyword(s):  
Shear Stress ◽  
Wall Shear Stress ◽  
Turbulent Flows ◽  
Patient Specific ◽  
4D Flow Mri ◽  
4D Flow ◽  
Local Flow ◽  
Flow Mri ◽  
Wall Shear

AbstractIn this study, we analyzed turbulent flows through a phantom (a 180$$^{\circ }$$ ∘ bend with narrowing) at peak systole and a patient-specific coarctation of the aorta (CoA), with a pulsating flow, using magnetic resonance imaging (MRI) and computational fluid dynamics (CFD). For MRI, a 4D-flow MRI is performed using a 3T scanner. For CFD, the standard $$k-\epsilon $$ k - ϵ , shear stress transport $$k-\omega $$ k - ω , and Reynolds stress (RSM) models are applied. A good agreement between measured and simulated velocity is obtained for the phantom, especially for CFD with RSM. The wall shear stress (WSS) shows significant differences between CFD and MRI in absolute values, due to the limited near-wall resolution of MRI. However, normalized WSS shows qualitatively very similar distributions of the local values between MRI and CFD. Finally, a direct comparison between in vivo 4D-flow MRI and CFD with the RSM turbulence model is performed in the CoA. MRI can properly identify regions with locally elevated or suppressed WSS. If the exact values of the WSS are necessary, CFD is the preferred method. For future applications, we recommend the use of the combined MRI/CFD method for analysis and evaluation of the local flow patterns and WSS in the aorta.

scholarly journals A novel MRI-based data fusion methodology for efficient, personalised, compliant simulations of aortic haemodynamics

2021 ◽  
Author(s):  
Catriona Stokes ◽  
Mirko Bonfanti ◽  
Zeyan Li ◽  
Jiang Xiong ◽  
Duanduan Chen ◽  
...  
Keyword(s):  
Data Fusion ◽  
Excellent Agreement ◽  
Computational Cost ◽  
Patient Specific ◽  
Cine Mri ◽  
Black Blood ◽  
4D Flow Mri ◽  
4D Flow ◽  
Wall Movement ◽  
Flow Mri

We present a novel, cost-efficient methodology to simulate aortic haemodynamics in a patient-specific, compliant aorta using an MRI data fusion process. Based on a previously-developed Moving Boundary Method, this technique circumvents the high computational cost and numerous structural modelling assumptions required by traditional Fluid-Structure Interaction techniques. Without the need for Computed Tomography (CT) data, the MRI images required to construct the simulation can be obtained during a single imaging session. Black Blood MR Angiography and 2D Cine-MRI data were used to reconstruct the luminal geometry and calibrate wall movement specifically to each region of the aorta. 4D-Flow MRI and non-invasive pressure measurements informed patient-specific inlet and outlet boundary conditions. Simulated wall movement closely matched 2D Cine-MRI measurements throughout the aorta, and physiological pressure and flow distributions in CFD were achieved within 3.3% of patient-specific targets. Excellent agreement with 4D-Flow MRI velocity data was observed. Conversely, a rigid-wall simulation under-predicted peak flow rate and systolic maximum velocities whilst predicting a mean Time-Averaged Wall Shear Stress (TAWSS) 13% higher than the compliant simulation. The excellent agreement observed between compliant simulation results and MRI is testament to the accuracy and efficiency of this MRI-based technique.

scholarly journals Patient-specific Hemodynamics of Severe Carotid Artery Stenosis Before and After Endarterectomy Examined by 4D Flow MRI

2019 ◽  
Vol 9 (1) ◽  
Author(s):  
Seungbin Ko ◽  
Jeesoo Lee ◽  
Simon Song ◽  
Doosang Kim ◽  
Sang Hyung Lee ◽  
...  
Keyword(s):  
Carotid Artery ◽  
Carotid Artery Stenosis ◽  
Artery Stenosis ◽  
Patch Repair ◽  
Patient Specific ◽  
4D Flow Mri ◽  
4D Flow ◽  
Spatiotemporal Resolution ◽  
Before And After ◽  
Flow Mri

AbstractCarotid endarterectomy (CEA) influences the carotid endoluminal anatomy, which results in hemodynamic changes before and after surgery. We investigated the hemodynamics of severe carotid artery stenosis before and after conventional endarterectomy with/without patch repair. An in vitro experiment utilizing carotid phantoms, which underwent a procedure that emulated CEA with/without the patch repair, was performed with a high-spatiotemporal resolution using 4D flow MRI. We evaluated an abnormal region of carotids, which consists of the normalized time-averaged wall shear stress (NTA|WSS|) and the oscillatory shear index (OSI), to account for continuous high-shear regions (high NTA|WSS| and low OSI) and chaotic low-shear regions, i.e., stenosis-prone regions (low NTA|WSS| and high OSI). The use of normalized hemodynamic parameters (e.g., NTA|WSS|) allowed comparison of diverse cases with different conditions of hemodynamics and vessel geometry. We observed that the stenosis-prone regions of the carotids with patches were noticeably larger than the corresponding regions in no-patch carotids. A large recirculating flow zone found in the stenosis-prone region of the internal carotid artery (ICA) of the postoperative carotids with patches partially blocks the flow path into ICA, and consequently the flow rate was not recovered after surgery unlike an expectation.

Sign in / Sign up

Export Citation Format

Share Document