Computational Fluid Dynamics modeling of contrast transport in basilar aneurysms following flow-altering surgeries

Comparison Between the Stereoscopic Virtual Reality Display System and Conventional Computed Tomography Workstation in the Diagnosis and Characterization of Cerebral Arteriovenous Malformations

It is difficult to accurately understand the angioarchitecture of cerebral arteriovenous malformations (CAVMs) before surgery using existing imaging methods. This study aimed to evaluate the ability of the stereoscopic virtual reality display system (SVRDS) to display the angioarchitecture of CAVMs by comparing its accuracy with that of the conventional computed tomography workstation (CCTW). Nineteen patients with CAVM confirmed on digital subtraction angiography (DSA) or during surgery were studied. Computed tomography angiography images in the SVRDS and CCTW were retrospectively analyzed by two experienced neuroradiologists using a double-blind method. Angioarchitectural parameters, such as the location and size of the nidus, type and number of the arterial feeders and draining veins, and draining pattern of the vessels, were recorded and compared. The diameter of the nidus ranged from 1.1 to 9 cm. Both CCTW and SVRDS correctly diagnosed the location of the nidus in 19 patients with CAVM. Among the 19 patients, 35 arterial feeders and 25 draining veins were confirmed on DSA and during surgery. With the DSA and intraoperative results as the gold standard bases, the CCTW misjudged one arterial feeder and one draining vein and missed three arterial feeders and two draining veins; meanwhile, the SVRDS missed only two arterial feeders. SVRDS had some advantages in displaying nidus, arterial branches, and draining veins of the CAVM compared with CCTW, as well as SVRDS could more intuitively display the overall angio-architectural spatial picture of CAVM.

Computational fluid dynamics-based virtual angiograms for the detection of flow stagnation in intracranial aneurysms

The goal of this study was to test if CFD-based virtual angiograms could be used to automatically discriminate between intracranial aneurysms (IAs) with and without flow stagnation. Time density curves (TDC) were extracted from patient digital subtraction angiography (DSA) image sequences by computing the average gray level intensity inside the aneurysm region and used to define injection profiles for each subject. Subject-specific 3D models were reconstructed from 3D rotational angiography (3DRA) and computational fluid dynamics (CFD) simulations were performed to simulate the blood flow inside IAs. Transport equations were solved numerically to simulate the dynamics of contrast injection into the parent arteries and IAs and then the contrast retention time (RET) was calculated. The importance of gravitational pooling of contrast agent within the aneurysm was evaluated by modeling contrast agent and blood as a mixture of two fluids with different densities and viscosities. Virtual angiograms can reproduce DSA sequences if the correct injection profile is used. RET can identify aneurysms with significant flow stagnation even when the injection profile is not known. Using a small sample of 14 IAs of which seven were previously classified as having flow stagnation, it was found that a threshold RET value of 0.46 s can successfully identify flow stagnation. CFD-based prediction of stagnation was in more than 90% agreement with independent visual DSA assessment of stagnation in a second sample of 34 IAs. While gravitational pooling prolonged contrast retention time it did not affect the predictive capabilities of RET. CFD-based virtual angiograms can detect flow stagnation in IAs and can be used to automatically identify aneurysms with flow stagnation even without including gravitational effects on contrast agents.

Recent developments in modeling, imaging, and monitoring of cardiovascular diseases using machine learning

Cardiovascular diseases are the leading cause of mortality, morbidity, and hospitalization around the world. Recent technological advances have facilitated analyzing, visualizing, and monitoring cardiovascular diseases using emerging computational fluid dynamics, blood flow imaging, and wearable sensing technologies. Yet, computational cost, limited spatiotemporal resolution, and obstacles for thorough data analysis have hindered the utility of such techniques to curb cardiovascular diseases. We herein discuss how leveraging machine learning techniques, and in particular deep learning methods, could overcome these limitations and offer promise for translation. We discuss the remarkable capacity of recently developed machine learning techniques to accelerate flow modeling, enhance the resolution while reduce the noise and scanning time of current blood flow imaging techniques, and accurate detection of cardiovascular diseases using a plethora of data collected by wearable sensors.

Challenges in Modeling Hemodynamics in Cerebral Aneurysms Related to Arteriovenous Malformations

PURPOSE

The significantly higher incidence of aneurysms in patients with arteriovenous malformations (AVMs) suggests a strong hemodynamic relationship between these lesions. The presence of an AVM alters hemodynamics in proximal vessels by drastically changing the distal resistance, thus affecting intra-aneurysmal flow. This study discusses the challenges associated with patient-specific modeling of aneurysms in the presence of AVMs.

METHODS

We explore how the presence of a generic distal AVM affects upstream aneurysms by examining the relationship between distal resistance and aneurysmal wall shear stress using physiologically realistic estimates for the influence of the AVM on hemodynamics. Using image-based computational models of aneurysms and surrounding vasculature, aneurysmal wall-shear stress is calculated for a range of distal resistances corresponding to the presence of AVMs of various sizes and compared with a control case representing the absence of an AVM.

RESULTS

In the patient cases considered, the alteration in aneurysmal wall shear stress due to the presence of an AVM is considerable, as much as 19 times the base case wall shear stress. Furthermore, the relationship between aneurysmal wall shear stress and distal resistance is shown to be highly geometry-dependent and nonlinear. In most cases, the range of physiologically realistic possibilities for AVM-related distal resistance are so large that patient-specific flow measurements are necessary for meaningful predictions of wall shear stress.

CONCLUSIONS

The presented work offers insight on the impact of distal AVMs on aneurysmal wall shear stress using physiologically realistic computational models. Patient-specific modeling of hemodynamics in aneurysms and associated AVMs has great potential for understanding lesion pathogenesis, surgical planning, and assessing the effect of treatment of one lesion relative to another. However, we show that modeling approaches cannot usually meaningfully quantify the impact of AVMs if based solely on imaging data from CT and X-ray angiography, currently used in clinical practice. Based on recent studies, it appears that 4D flow MRI is one promising approach to obtaining meaningful patient-specific flow boundary conditions that improve modeling fidelity.

Comparison of Newtonian and Non-newtonian Fluid Models in Blood Flow Simulation in Patients With Intracranial Arterial Stenosis

Background

Newtonian fluid model has been commonly applied in simulating cerebral blood flow in intracranial atherosclerotic stenosis (ICAS) cases using computational fluid dynamics (CFD) modeling, while blood is a shear-thinning non-Newtonian fluid. We aimed to investigate the differences of cerebral hemodynamic metrics quantified in CFD models built with Newtonian and non-Newtonian fluid assumptions, in patients with ICAS.

Methods

We built a virtual artery model with an eccentric 75% stenosis and performed static CFD simulation. We also constructed CFD models in three patients with ICAS of different severities in the luminal stenosis. We performed static simulations on these models with Newtonian and two non-Newtonian (Casson and Carreau-Yasuda) fluid models. We also performed transient simulations on another patient-specific model. We measured translesional pressure ratio (PR) and wall shear stress (WSS) values in all CFD models, to reflect the changes in pressure and WSS across a stenotic lesion. In all the simulations, we compared the PR and WSS values in CFD models derived with Newtonian, Casson, and Carreau-Yasuda fluid assumptions.

Results

In all the static and transient simulations, the Newtonian/non-Newtonian difference on PR value was negligible. As to WSS, in static models (virtual and patient-specific), the rheological difference was not obvious in areas with high WSS, but observable in low WSS areas. In the transient model, the rheological difference of WSS areas with low WSS was enhanced, especially during diastolic period.

Conclusion

Newtonian fluid model could be applicable for PR calculation, but caution needs to be taken when using the Newtonian assumption in simulating WSS especially in severe ICAS cases.

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

BACKGROUND AND OBJECTIVE

Time resolved three-dimensional phase contrast magnetic resonance imaging (4D-Flow MRI) has been used to non-invasively measure blood velocities in the human vascular system. However, issues such as low spatio-temporal resolution, acquisition noise, velocity aliasing, and phase-offset artifacts have hampered its clinical application. In this research, we developed a purely data-driven method for super-resolution and denoising of 4D-Flow MRI.

METHODS

The flow velocities, pressure, and the MRI image magnitude are modeled as a patient-specific deep neural net (DNN). For training, 4D-Flow MRI images in the complex Cartesian space are used to impose data-fidelity. Physics of fluid flow is imposed through regularization. Creative loss function terms have been introduced to handle noise and super-resolution. The trained patient-specific DNN can be sampled to generate noise-free high-resolution flow images. The proposed method has been implemented using the TensorFlow DNN library and tested on numerical phantoms and validated in-vitro using high-resolution particle image velocitmetry (PIV) and 4D-Flow MRI experiments on transparent models subjected to pulsatile flow conditions.

RESULTS

In case of numerical phantoms, we were able to increase spatial resolution by a factor of 100 and temporal resolution by a factor of 5 compared to the simulated 4D-Flow MRI. There is an order of magnitude reduction of velocity normalized root mean square error (vNRMSE). In case of the in-vitro validation tests with PIV as reference, there is similar improvement in spatio-temporal resolution. Although the vNRMSE is reduced by 50%, the method is unable to negate a systematic bias with respect to the reference PIV that is introduced by the 4D-Flow MRI measurement.

CONCLUSIONS

This work has demonstrated the feasibility of using the readily available machinery of deep learning to enhance 4D-Flow MRI using a purely data-driven method. Unlike current state-of-the-art methods, the proposed method is agnostic to geometry and boundary conditions and therefore eliminates the need for tedious tasks such as accurate image segmentation for geometry, image registration, and estimation of boundary flow conditions. Arbitrary regions of interest can be selected for processing. This work will lead to user-friendly analysis tools that will enable quantitative hemodynamic analysis of vascular diseases in a clinical setting.

Towards multi-modal data fusion for super-resolution and denoising of 4D-Flow MRI

4D-Flow magnetic resonance imaging (MRI) has enabled in vivo time-resolved measurement of three-dimensional blood flow velocities in the human vascular system. However, its clinical use has been hampered by two main issues, namely, low spatio-temporal resolution and acquisition noise. While patient-specific computational fluid dynamics (CFD) simulations can address the resolution and noise issues, its fidelity is impacted by accuracy of estimation of boundary conditions, model parameters, vascular geometry, and flow model assumptions. In this paper a scheme to address limitations of both modalities through data-fusion is presented. The solutions of the patient-specific CFD simulation are characterized using proper orthogonal decomposition (POD). Next, a process of projecting the 4D-Flow MRI data onto the POD basis and projection coefficient mapping using generalized dynamic mode decomposition (DMD) enables simultaneous super-resolution and denoising of 4D-Flow MRI. The method has been tested using numerical phantoms derived from patient-specific aneurysmal geometries and applied to in vivo 4D-Flow MRI data.

Fusiform aneurysms of the vertebrobasilar complex: a single-center series

BACKGROUND

Fusiform vertebrobasilar aneurysms (FVBAs) may exhibit a disastrous clinical course. Due to their rare occurrence, evidence concerning optimal management is lackluster.

OBJECTIVE

To describe the epidemiology, clinical features and treatment outcomes of a consecutive series of patients admitted to our institution.

METHODS

We retrospectively evaluated patient charts with respect to clinical presentation, treatment procedures, and the outcomes of all patients diagnosed with an FVBA, which were seen at our institution between March 2006 and February 2017.

RESULTS

Forty-five consecutive patients were analyzed. Follow-up was available for 39 patients (86.7%) with a median duration of 28.8 months. Seventeen patients (37.7%) were asymptomatic, 14 patients (31.1%) presented with brainstem ischemia, 8 patients (17.8%) with supratentorial ischemia, and 3 (6.7%) patients with brain stem compression. Aneurysm rupture occurred in 3 patients upon presentation (6.7%). Initially, 19 patients (42.2%) were significantly disabled with Modified Rankin Scale (mRS) scores ≥ 3. Twelve patients (26.7%) underwent invasive treatment: endovascular therapy in 9 cases and surgical treatment in 3 cases. Thirty-three patients received conservative treatment. During follow-up, 6 events (66.7%) of severe disability or death (mRS 4-6) occurred in the endovascular group versus 1 event (33%) in the surgical group versus 19 events (63.3%) among conservatively treated aneurysms. Deterioration was significantly more frequent in patients with symptomatic aneurysms (p = 0.030).

CONCLUSION

Patients harboring an FVBA frequently present with disabling symptoms caused by various pathomechanisms. The natural history is aggressive, mostly for initially symptomatic aneurysms, and periprocedural morbidity of surgical or endovascular treatment remains substantial.

Computational Hemodynamic Modeling of Arterial Aneurysms: A Mini-Review

Arterial aneurysms are pathological dilations of blood vessels, which can be of clinical concern due to thrombosis, dissection, or rupture. Aneurysms can form throughout the arterial system, including intracranial, thoracic, abdominal, visceral, peripheral, or coronary arteries. Currently, aneurysm diameter and expansion rates are the most commonly used metrics to assess rupture risk. Surgical or endovascular interventions are clinical treatment options, but are invasive and associated with risk for the patient. For aneurysms in locations where thrombosis is the primary concern, diameter is also used to determine the level of therapeutic anticoagulation, a treatment that increases the possibility of internal bleeding. Since simple diameter is often insufficient to reliably determine rupture and thrombosis risk, computational hemodynamic simulations are being developed to help assess when an intervention is warranted. Created from subject-specific data, computational models have the potential to be used to predict growth, dissection, rupture, and thrombus-formation risk based on hemodynamic parameters, including wall shear stress, oscillatory shear index, residence time, and anomalous blood flow patterns. Generally, endothelial damage and flow stagnation within aneurysms can lead to coagulation, inflammation, and the release of proteases, which alter extracellular matrix composition, increasing risk of rupture. In this review, we highlight recent work that investigates aneurysm geometry, model parameter assumptions, and other specific considerations that influence computational aneurysm simulations. By highlighting modeling validation and verification approaches, we hope to inspire future computational efforts aimed at improving our understanding of aneurysm pathology and treatment risk stratification.

Hemodynamics of Cerebral Aneurysms: Connecting Medical Imaging and Biomechanical Analysis

In the last two decades, numerous studies have conducted patient-specific computations of blood flow dynamics in cerebral aneurysms and reported correlations between various hemodynamic metrics and aneurysmal disease progression or treatment outcomes. Nevertheless, intra-aneurysmal flow analysis has not been adopted in current clinical practice, and hemodynamic factors usually are not considered in clinical decision making. This review presents the state of the art in cerebral aneurysm imaging and image-based modeling, discussing the advantages and limitations of each approach and focusing on the translational value of hemodynamic analysis. Combining imaging and modeling data obtained from different flow modalities can improve the accuracy and fidelity of resulting velocity fields and flow-derived factors that are thought to affect aneurysmal disease progression. It is expected that predictive models utilizing hemodynamic factors in combination with patient medical history and morphological data will outperform current risk scores and treatment guidelines. Possible future directions include novel approaches enabling data assimilation and multimodality analysis of cerebral aneurysm hemodynamics.

Standardized Evaluation of Cerebral Arteriovenous Malformations Using Flow Distribution Network Graphs and Dual-venc 4D Flow MRI

BACKGROUND

Cerebral arteriovenous malformations (AVMs) are pathological connections between arteries and veins. Dual-venc 4D flow MRI, an extended 4D flow MRI method with improved velocity dynamic range, provides time-resolved 3D cerebral hemodynamics.

PURPOSE

To optimize dual-venc 4D flow imaging parameters for AVM; to assess the relationship between spatial resolution, acceleration, and flow quantification accuracy; and to introduce and apply the flow distribution network graph (FDNG) paradigm for storing and analyzing complex neurovascular 4D flow data.

STUDY TYPE

Retrospective cohort study.

SUBJECTS/PHANTOM

Scans were performed in a specialized flow phantom: 26 healthy subjects (age 41 ± 17 years) and five AVM patients (age 27-68 years).

FIELD STRENGTH/SEQUENCE

Dual-venc 4D flow with varying spatial resolution and acceleration factors were performed at 3T field strength.

ASSESSMENT

Quantification accuracy was assessed in vitro by direct comparison to measured flow. FDNGs were used to quantify and compare flow, peak velocity (PV), and pulsatility index (PI) between healthy controls with various Circle of Willis (CoW) anatomy and AVM patients.

STATISTICAL TESTS

In vitro measurements were compared to ground truth with Student's t-test. In vivo groups were compared with Wilcoxon rank-sum test and Kruskal-Wallis test.

RESULTS

Flow was overestimated in all in vitro experiments, by an average 7.1 ± 1.4% for all measurement conditions. Error in flow measurement was significantly correlated with number of voxels across the channel (P = 3.11 × 10^-28 ) but not with acceleration factor (P = 0.74). For the venous-arterial PV and PI ratios, a significant difference was found between AVM nidal and extranidal circulation (P = 0.008 and 0.05, respectively), and between AVM nidal and healthy control circulation (P = 0.005 and 0.003, respectively).

DATA CONCLUSION

Dual-venc 4D flow MRI and standardized FDNG analysis might be feasible in clinical applications. Venous-arterial ratios of PV and PI are proposed as network-based biomarkers characterizing AVM nidal hemodynamics.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;50:1718-1730.

Semi-automated analysis of 4D flow MRI to assess the hemodynamic impact of intracranial atherosclerotic disease

PURPOSE

This study evaluated the feasibility of using 4D flow MRI and a semi-automated analysis tool to assess the hemodynamic impact of intracranial atherosclerotic disease (ICAD). The ICAD impact was investigated by evaluating pressure drop (PD) at the atherosclerotic stenosis and changes in cerebral blood flow distribution in patients compared to healthy controls.

METHODS

Dual-venc 4D flow MRI was acquired in 25 healthy volunteers and 16 ICAD patients (ICA, N = 3; MCA, N = 13) with mild (<50%), moderate (50-69%), or severe (>70%) intracranial stenosis. A semi-automated analysis tool was developed to quantify velocity and flow from 4D flow MRI and to evaluate cerebral blood flow redistribution. PD at stenosis was estimated using the Bernoulli equation. The PD calculation was examined by an in vitro phantom study against flow simulations.

RESULTS

Flow analysis in controls indicated symmetry in blood flow rate (FR) and peak velocity (PV) between the brain hemispheres. For patients, PV in the affected hemisphere was significantly (65%) higher than the normal side (P = 0.002). However, FR to both hemispheres of the brain was the same. The PD depicted significant correlation with PV asymmetry in patients (ρ = 0.67 and P = 0.02), and it was significantly higher for severe compared to moderate stenosis (3.73 vs. 2.30 mm Hg, P = 0.02).

CONCLUSION

4D flow MRI quantification enables assessment of the hemodynamic impact of ICAD. The significant difference of the PD between patients with severe and moderate stenosis and its correlation with PV asymmetry suggest that PD may be a pertinent hemodynamic biomarker to evaluate ICAD.

Mapping the Transport Kinetics of Molecules and Particles in Idealized Intracranial Side Aneurysms

Intracranial side aneurysms (IA) are pathological blood-filled bulges in cerebral blood vessels. Unlike healthy blood vessels where mass transport is dominated by convection, both diffusion and convection can play an active role in aneurysm sites. Here, we study via dye washout experiments and numerical simulations, the transport characteristics of particles (1 micron) and small molecules (300 Da) into simplified side aneurysms models following bolus injection. Time-lapse fluorescent microscopy imaging performed in our idealized aneurysm models showed that the parent artery geometry (located on the inner vs. outer curvature) as well as the aneurysm aspect ratio (AR) affect the washout kinetics while the pulsatile nature of the flow, maintained within the physiological range, carries only a minor effect. Importantly, in the absence of effective diffusion, particles that are located on slow streamlines linger within the aneurysm cavity, a phenomenon that could be of importance in deposition of cells and nano/micro-particles within aneurysms. Altogether, mass transport studies may provide valuable insights for better understanding of aneurysm pathophysiology as well as for the design of new diagnostic and theranostic nano-medicines.

Pressure and Flow Rate Changes During Contrast Injections in Cerebral Angiography: Correlation to Reflux Length

Cerebral angiography involves the antegrade injection of contrast media through a catheter into the vasculature to visualize the region of interest under X-ray imaging. Depending on the injection and blood flow parameters, the bolus of contrast can propagate in the upstream direction and proximal to the catheter tip, at which point contrast is said to have refluxed. In this in vitro study, we investigate the relationship of fundamental hemodynamic variables to this phenomenon. Contrast injections were carried out under steady and pulsatile flow using various vessel diameters, catheter sizes, working fluid flow rates, and injection rates. The distance from the catheter tip to the proximal edge of the contrast bolus, called reflux length, was measured on the angiograms; the relation of this reflux length to different hemodynamic parameters was evaluated. Results show that contrast reflux occurs when the pressure distal to the catheter tip increases to be greater than the pressure proximal to the catheter tip. The ratio of this pressure difference to the baseline flow rate, called reflux resistance here, was linearly correlated to the normalized reflux length (reflux length/vessel diameter). Further, the ratio of blood flow to contrast fluid momentums, called the Craya-Curtet number, was correlated to the normalized reflux length via a sigmoid function. A sigmoid function was also found to be representative of the relationship between the ratio of the Reynolds numbers of blood flow to contrast and the normalized reflux length. As described by previous reports, catheter based contrast injections cause substantial increases in local flow and pressure. Contrast reflux should generally be avoided during standard antegrade angiography. Our study shows two specific correlations between contrast reflux length and baseline and intra-injection parameters that have not been published previously. Further studies need to be conducted to fully characterize the phenomena and to extract reliable indicators of clinical utility. Parameters relevant to cerebral angiography are studied here, but the essential principles are applicable to all angiographic procedures involving antegrade catheter injections.

Merging computational fluid dynamics and 4D Flow MRI using proper orthogonal decomposition and ridge regression

Time resolved phase-contrast magnetic resonance imaging 4D-PCMR (also called 4D Flow MRI) data while capable of non-invasively measuring blood velocities, can be affected by acquisition noise, flow artifacts, and resolution limits. In this paper, we present a novel method for merging 4D Flow MRI with computational fluid dynamics (CFD) to address these limitations and to reconstruct de-noised, divergence-free high-resolution flow-fields. Proper orthogonal decomposition (POD) is used to construct the orthonormal basis of the local sampling of the space of all possible solutions to the flow equations both at the low-resolution level of the 4D Flow MRI grid and the high-level resolution of the CFD mesh. Low-resolution, de-noised flow is obtained by projecting in vivo 4D Flow MRI data onto the low-resolution basis vectors. Ridge regression is then used to reconstruct high-resolution de-noised divergence-free solution. The effects of 4D Flow MRI grid resolution, and noise levels on the resulting velocity fields are further investigated. A numerical phantom of the flow through a cerebral aneurysm was used to compare the results obtained using the POD method with those obtained with the state-of-the-art de-noising methods. At the 4D Flow MRI grid resolution, the POD method was shown to preserve the small flow structures better than the other methods, while eliminating noise. Furthermore, the method was shown to successfully reconstruct details at the CFD mesh resolution not discernible at the 4D Flow MRI grid resolution. This method will improve the accuracy of the clinically relevant flow-derived parameters, such as pressure gradients and wall shear stresses, computed from in vivo 4D Flow MRI data.