Toward a physiological model of vascular wall vibrations in the arteriovenous fistula.

The mechanism behind hemodialysis arteriovenous fistula (AVF) failure remains poorly understood, despite previous efforts to correlate altered hemodynamics with vascular remodeling. We have recently demonstrated that transitional flow induces high-frequency vibrations in the AVF wall, albeit with a simplified model. This study addresses the key limitations of our original fluid-structure interaction (FSI) approach, aiming to evaluate the vibration response using a more realistic model. A 3D AVF geometry was generated from contrast-free MRI and high-fidelity FSI simulations were performed. Patient-specific inflow and pressure were incorporated, and a three-term Mooney-Rivlin model was fitted using experimental data. The viscoelastic effect of perivascular tissue was modeled with Robin boundary conditions. Prescribing pulsatile inflow and pressure resulted in a substantial increase in vein displacement ( + 400 %) and strain ( + 317 %), with a higher maximum spectral frequency becoming visible above -42 dB (from 200 to 500 Hz). Transitioning from Saint Venant-Kirchhoff to Mooney-Rivlin model led to displacement amplitudes exceeding 10 micrometers and had a substantial impact on strain ( + 116 %). Robin boundary conditions significantly damped high-frequency displacement ( - 60 %). Incorporating venous tissue properties increased vibrations by 91%, extending up to 700 Hz, with a maximum strain of 0.158. Notably, our results show localized, high levels of vibration at the inner curvature of the vein, a site known for experiencing pronounced remodeling. Our findings, consistent with experimental and clinical reports of bruits and thrills, underscore the significance of incorporating physiologically plausible modeling approaches to investigate the role of wall vibrations in AVF remodeling and failure.

Impact of left atrial wall motion assumptions in fluid simulations on proposed predictors of thrombus formation.

Atrial fibrillation (AF) poses a significant risk of stroke due to thrombus formation, which primarily occurs in the left atrial appendage (LAA). Medical image-based computational fluid dynamics (CFD) simulations can provide valuable insight into patient-specific hemodynamics and could potentially enhance personalized assessment of thrombus risk. However, the importance of accurately representing the left atrial (LA) wall dynamics has not been fully resolved. In this study, we compared four modeling scenarios; rigid walls, a generic wall motion based on a reference motion, a semi-generic wall motion based on patient-specific motion, and patient-specific wall motion based on medical images. We considered a LA geometry acquired from 4D computed tomography during AF, systematically performed convergence tests to assess the numerical accuracy of our solution strategy, and quantified the differences between the four approaches. The results revealed that wall motion had no discernible impact on LA cavity hemodynamics, nor on the markers that indicate thrombus formation. However, the flow patterns within the LAA deviated significantly in the rigid model, indicating that the assumption of rigid walls may lead to errors in the estimated risk factors. In contrast, the generic, semi-generic, and patient-specific cases were qualitatively similar. The results highlight the crucial role of wall motion on hemodynamics and predictors of thrombus formation, and also demonstrate the potential of using a generic motion model as a surrogate for the more complex patient-specific motion. While the present study considered a single case, the employed CFD framework is entirely open-source and designed for adaptability, allowing for integration of additional models and generic motions.

On the importance of fundamental computational fluid dynamics toward a robust and reliable model of left atrial flows.

Computational fluid dynamics (CFD) studies of left atrial flows have reached a sophisticated level, for example, revealing plausible relationships between hemodynamics and stresses with atrial fibrillation. However, little focus has been on fundamental fluid modeling of LA flows. The purpose of this study was to investigate the spatiotemporal convergence, along with the differences between high- (HR) versus normal-resolution/accuracy (NR) solution strategies, respectively. Rigid wall CFD simulations were conducted on 12 patient-specific left atrial geometries obtained from computed tomography scans, utilizing a second-order accurate and space/time-centered solver. The convergence studies showed an average variability of around 30% and 55% for time averaged wall shear stress (WSS), oscillatory shear index (OSI), relative residence time (RRT), and endothelial cell activation potential (ECAP), even between intermediate spatial and temporal resolutions, in the left atrium (LA) and left atrial appendage (LAA), respectively. The comparison between HR and NR simulations showed good correlation in the LA for WSS, RRT, and ECAP ( R 2 > .9 ), but not for OSI ( R 2 = .63 ). However, there were poor correlations in the LAA especially for OSI, RRT, and ECAP ( R 2 = .55, .63, and .61, respectively), except for WSS ( R 2 = .81 ). The errors are comparable to differences previously reported with disease correlations. To robustly predict atrial hemodynamics and stresses, numerical resolutions of 10 M elements (i.e., Δ x = ∼ .5 mm) and 10 k time-steps per cycle seem necessary (i.e., one order of magnitude higher than normally used in both space and time). In conclusion, attention to fundamental numerical aspects is essential toward establishing a plausible, robust, and reliable model of LA flows.

Flow-induced high frequency vascular wall vibrations in an arteriovenous fistula: a specific stimulus for stenosis development?

Hemodialysis is the lifeline for nearly three million end stage renal disease patients worldwide. Native arteriovenous fistula (AVF) is the preferred vascular access, but 40% fail within 1 year. We recently demonstrated that AVFs harbour transitional flows and the goal of the present study was to investigate whether the associated high-frequency pressure fluctuations could promote vibrations within the vascular wall. We acquired MRI images and flow rates immediately after surgery in one patient and generated a 3D patient-specific model. High-fidelity fluid structure interaction simulations revealed the presence of wall vibrations in distinct frequency bands up to 200 Hz and amplitude of 200 µm. A sensitivity analysis to assess the impact of flow rates, and vascular wall stiffness and thickness, changes that typically occur during AVF maturation, confirmed the robustness of the results. Interestingly, the vibrations were always predominant at the anastomosis floor and on the inner venous side, which correlates with typical stenotic regions. As studies seeking to correlate aberrant stresses and vascular remodelling have been largely inconclusive, the focal colocalization between vibrations and stenosis may suggest an unknown mechanobiological process between high-frequency mechanical stresses within the vascular wall and adverse vascular remodelling.

Understanding intracranial aneurysm sounds via high-fidelity fluid-structure-interaction modelling.

BACKGROUND: Since the 1960s, the origins of intracranial aneurysm bruits and musical murmurs have been debated, with proposed mechanisms ranging from self-excitation (i.e., resonance) by stable pulsatile flow, to vibration caused by unstable (laminar vortex shedding or turbulent) flow. This knowledge gap has impeded the use of intracranial sounds a marker of aneurysm remodelling or rupture risk. New computational techniques now allow us to model these phenomena. METHODS: We performed high-fidelity fluid-structure interaction simulations capable of understanding the magnitude and mechanisms of such flow-induced vibrations, under pulsatile flow conditions. Six cases from a previous cohort were used. RESULTS: In five cases, underlying flow instabilities present as broad-band, random vibrations, consistent with previously-described bruits, while the sac also exhibits resonance, rocking back and forth in different planes of motion, consistent with previously described musical murmurs. Both types of vibration have amplitudes in the range of 0.1 to 1 µm. The murmurs extend into diastole, after the underlying flow instability has dissipated, and do not exhibit the characteristic repeating frequency harmonics of previously hypothesized vortex-shedding mechanisms. The remaining case with stable pulsatile flow does not vibrate. Spectrograms of the simulated vibrations are consistent with previously reported microphone and Doppler ultrasound recordings. CONCLUSIONS: Our results provide a plausible explanation for distinct intracranial aneurysm sounds and characterize the mechanical environment of a vibrating aneurysm wall. Future work should aim to quantify the deleterious effects of these overlooked stimuli on the vascular wall, to determine which changes to the wall makeup are associated with vibration.

Brain aneurysms are weak areas of an artery in the brain that form a bulge. Most aneurysms do not rupture, but when they do, most patients die or are severely disabled. Unruptured aneurysms are often found by chance, but the risk of complications from treating them can be higher than the risk of rupture. Aneurysms are known to produce sounds, but the connection between the sounds and the resulting vibrations that may further weaken the artery wall are poorly understood. In this study, we modelled the possible vibration of aneurysm walls during turbulent blood flow. The frequency patterns of the vibrations were consistent with sound recordings obtained from patients. Our findings could enable potentially problematic aneurysms to be more easily identified in the future.

A verified and validated moving domain computational fluid dynamics solver with applications to cardiovascular flows.

Computational fluid dynamics (CFD) in combination with patient-specific medical images has been used to correlate flow phenotypes with disease initiation, progression and outcome, in search of a prospective clinical tool. A large number of CFD software packages are available, but are typically based on rigid domains and low-order finite volume methods, and are often implemented in massive low-level C++ libraries. Furthermore, only a handful of solvers have been appropriately verified and validated for their intended use. Our goal was to develop, verify and validate an open-source CFD solver for moving domains, with applications to cardiovascular flows. The solver is an extension of the CFD solver Oasis, which is based on the finite element method and implemented using the FEniCS open source framework. The new solver, named OasisMove, extends Oasis by expressing the Navier-Stokes equations in the arbitrary Lagrangian-Eulerian formulation, which is suitable for handling moving domains. For code verification we used the method of manufactured solutions for a moving 2D vortex problem, and for validation we compared our results against existing high-resolution simulations and laboratory experiments for two moving domain problems of varying complexity. Verification results showed that the L 2 error followed the theoretical convergence rates. The temporal accuracy was second-order, while the spatial accuracy was second- and third-order using ℙ 1 / ℙ 1 and ℙ 2 / ℙ 1 finite elements, respectively. Validation results showed good agreement with existing benchmark results, by reproducing lift and drag coefficients with less than 1% error, and demonstrating the solver's ability to capture vortex patterns in transitional and turbulent-like flow regimes. In conclusion, we have shown that OasisMove is an open-source, accurate and reliable solver for cardiovascular flows in moving domains.

Onset and nature of flow-induced vibrations in cerebral aneurysms via fluid-structure interaction simulations.

Clinical, experimental, and recent computational studies have demonstrated the presence of wall vibrations in cerebral aneurysms, thought to be induced by blood flow instability. These vibrations could induce irregular, high-rate deformation of the aneurysm wall, and potentially disrupt regular cell behavior and promote deleterious wall remodeling. In order to elucidate, for the first time, the onset and nature of such flow-induced vibrations, in this study we imposed a linearly increasing flow rate on high-fidelity fluid-structure interaction models of three anatomically realistic aneurysm geometries. Prominent narrow-band vibrations in the range of 100-500 Hz were found in two out of the three aneurysm geometries tested, while the case that did not exhibit flow instability did not vibrate. Aneurysm vibrations consisted mostly of fundamental modes of the entire aneurysm sac, with the vibrations exhibiting more frequency content at higher frequencies than the flow instabilities driving those vibrations. The largest vibrations occurred in the case which exhibited strongly banded fluid frequency content, and the vibration amplitude was highest when the strongest fluid frequency band was an integer multiple of one of the natural frequencies of the aneurysm sac. Lower levels of vibration occurred in the case which exhibited turbulent-like flow with no distinct frequency bands. The current study provides a plausible mechanistic explanation for the high-frequency sounds observed in cerebral aneurysms, and suggests that narrow-band (vortex-shedding type) flow might stimulate the wall more, or at least at lower flow rates, than broad-band, turbulent-like flow.

High-fidelity fluid structure interaction simulations of turbulent-like aneurysm flows reveals high-frequency narrowband wall vibrations: A stimulus of mechanobiological relevance?

Recent high-fidelity/resolution computational fluid dynamics simulations of intracranial aneurysm hemodynamics have revealed turbulent-like flows. We hypothesized that the associated high-frequency pressure fluctuations could promote aneurysm wall vibrations. We performed fully coupled high-fidelity transient fluid structure interaction simulations between the blood flow and compliant aneurysm sac wall taking 5,000 time steps per second using a 3D patient-specific model previously shown to harbour turbulent-like flow. Our results show that the flow velocity contained fluctuations with a smooth and continuously decaying energy up to ∼160Hz, and fluctuating pressures with characteristic frequency peaks at approximately 30, 130 and 210Hz. There was a strong two-way coupling between the pressure and the wall deformation, for which the frequency spectrum showed similar characteristics, but with a narrow band peak at ∼120Hz with large regional differences in amplitude up to 80µm. The physics of the flow is broadly consistent with clinical reports of turbulent-like flows, while the physics of the wall is consistent with reports of spectral peaks in aneurysm patients. As many aneurysms are known to harbour turbulent-like flows, wall vibrations could be a widespread phenomenon. Finally, since aneurysms are vascular pathologies by definition and many/most aneurysms do not have endothelial cells but still display a focal remodeling, we hypothesize that vibrations and stresses within the wall itself might play a role in the mechanobiological processes of vessel wall pathology.

Automated landmarking of bends in vascular structures: a comparative study with application to the internal carotid artery.

Automated tools for landmarking the internal carotid artery (ICA) bends have the potential for efficient and objective medical image-based morphometric analysis. The two existing algorithms rely on numerical approximations of curvature and torsion of the centerline. However, input parameters, original source code, comparability, and robustness of the algorithms remain unknown. To address the former two, we have re-implemented the algorithms, followed by sensitivity analyses. Of the input parameters, the centerline smoothing had the least impact resulting in 6-7 bends, which is anatomically realistic. In contrast, centerline resolution showed to completely over- and underestimated the number of bends varying from 3 to 33. Applying the algorithms to the same cohort revealed a variability that makes comparison of results between previous studies questionable. Assessment of robustness revealed how one algorithm is vulnerable to model smoothness and noise, but conceptually independent of application. In contrast, the other algorithm is robust and consistent, but with limited general applicability. In conclusion, both algorithms are equally valid albeit they produce vastly different results. We have provided a well-documented open-source implementation of the algorithms. Finally, we have successfully performed this study on the ICA, but application to other vascular regions should be performed with caution.

Computed Poststenotic Flow Instabilities Correlate Phenotypically With Vibrations Measured Using Laser Doppler Vibrometry: Perspectives for a Promising In Vivo Device for Early Detection of Moderate and Severe Carotid Stenosis.

Early detection of asymptomatic carotid stenosis is crucial for treatment planning in the prevention of ischemic stroke. Auscultation, the current first-line screening methodology, comes with severe limitations that create urge for novel and robust techniques. Laser Doppler vibrometer (LDV) is a promising tool for inferring carotid stenosis by measuring stenosis-induced vibrations. The goal of the current study was to evaluate the feasibility of LDV for carotid stenosis detection. LDV measurements on a carotid phantom were used to validate our previously verified high-resolution computational fluid dynamics methodology, which was used to evaluate the impact of flowrate, flow split, and stenosis severity on the poststenotic intensity of flow instabilities (IFI). We evaluated sensitivity, specificity, and accuracy of using IFI for stenoses detection. Linear regression analyses showed that computationally derived pressure fluctuations correlated (R2 = 0.98) with LDV measurements of stenosis-induced vibrations. The flowrate of stenosed vessels correlated (R2 = 0.90) with the presence of poststenotic instabilities. Receiver operating characteristic analyses of power spectra revealed that the most relevant frequency bands for the detection of moderate (56-76%) and severe (86-96%) stenoses were 80-200 Hz and 0-40 Hz, respectively. Moderate stenosis was identified with sensitivity and specificity of 90%; values decreased to 70% for severe stenosis. The use of LDV as screening tool for asymptomatic stenosis can potentially provide improved accuracy of current screening methodologies for early detection. The applicability of this promising device for mass screening is currently being evaluated clinically.

A framework for automated and objective modification of tubular structures: Application to the internal carotid artery.

Patient-specific medical image-based computational fluid dynamics has been widely used to reveal fundamental insight into mechanisms of cardiovascular disease, for instance, correlating morphology to adverse vascular remodeling. However, segmentation of medical images is laborious, error-prone, and a bottleneck in the development of large databases that are needed to capture the natural variability in morphology. Instead, idealized models, where morphological features are parameterized, have been used to investigate the correlation with flow features, but at the cost of limited understanding of the complexity of cardiovascular flows. To combine the advantages of both approaches, we developed a tool that preserves the patient-specificness inherent in medical images while allowing for parametric alteration of the morphology. In our open-source framework morphMan we convert the segmented surface to a Voronoi diagram, modify the diagram to change the morphological features of interest, and then convert back to a new surface. In this paper, we present algorithms for modifying bifurcation angles, location of branches, cross-sectional area, vessel curvature, shape of bends, and surface roughness. We show qualitative and quantitative validation of the algorithms, performing with an accuracy exceeding 97% in general, and proof-of-concept on combining the tool with computational fluid dynamics. By combining morphMan with appropriate clinical measurements, one could explore the morphological parameter space and resulting hemodynamic response using only a handful of segmented surfaces, effectively minimizing the main bottleneck in image-based computational fluid dynamics.

Automated and objective removal of bifurcation aneurysms: Incremental improvements, and validation against healthy controls.

Abnormal hemodynamic stresses are thought to correlate with aneurysm initiation, growth, and rupture. We have previously investigated the role of wall shear stress (WSS) and WSS gradients (WSSG) in search for a mechanistic link to formation of sidewall aneurysms using an automated and objective tool for aneurysm removal and arterial reconstruction in combination with computational fluid dynamics (CFD). However, we warned against the use of the tool for bifurcation type aneurysms because of a potential unrealistic reconstruction of the apex. We hypothesized that inclusion of additional morphological features from the surrounding vasculature could overcome these constraints. We extended the previously published method for removal and reconstruction of the bifurcation vasculature based on diverging and converging points of the parent and daughter artery centerlines, to also include two new centerlines between the daughter vessels, one of them passed through the bifurcation center. Validation was performed by comparing the efficacy of the two algorithms, using ten healthy models of the internal carotid artery terminus as ground truth. Qualitative results showed that the bifurcation apexes became smoother relative to the original algorithm; more consistent with the reference models. This was reflected quantitatively by a reduced maximum distance between the reference and reconstructed surfaces, although not statistically significant. Furthermore, the modified algorithm also quantitatively improved CFD derived WSS and WSSG, especially the latter. In conclusion, the modified algorithm does not perfectly reconstruct the bifurcation apex, but provides an incremental improvement, especially important for the derived hemodynamic metrics of interest in vascular pathobiology.

Time-Lapsing Perfusion: Proof of Concept of a Novel Method to Study Drug Delivery in Whole Organs.

Perfusion is one of the most important processes maintaining organ health. From a computational perspective, however, perfusion is among the least-studied physiological processes of the heart. The recent development of novel nanoparticle-based targeted cardiac therapy calls for novel simulation methods that can provide insights into the distribution patterns of therapeutic agents within the heart tissue. Additionally, resolving the distribution patterns of perfusion is crucial for gaining a full understanding of the long-term impacts of cardiovascular diseases that can lead to adverse remodeling such as myocardial ischemia and heart failure. In this study, we have developed and used a, to our knowledge, novel particle-tracking-based method to simulate the perfusion-mediated distribution of nanoparticles or other solutes. To model blood flow through perfused tissue, we follow the approach of others and treat the tissue as a porous medium in a continuum model. Classically, solutes are modeled using reaction-advection-diffusion kinetics. However, because of the discrepancy of scales between advection and diffusion in blood vessels, this method becomes practically numerically unstable. Instead, we track a bolus of solutes or nanoparticles using particle tracking based purely on advection in arteries. In capillaries, we employ diffusion kinetics, using an effective diffusion coefficient to mimic capillary blood flow. We first demonstrate the numerical validity and computational efficiency of this method on a two-dimensional benchmark problem. Finally, we demonstrate how the method is used to visualize perfusion patterns of a healthy and ischemic human left ventricle geometry. The efficiency of the method allows for nanoparticle tracking over multiple cardiac cycles using a conventional laptop, providing a framework for the simulation of experimentally relevant timeframes to advance preclinical research.

Multiple Aneurysms AnaTomy CHallenge 2018 (MATCH)-phase II: rupture risk assessment.

PURPOSE: Assessing the rupture probability of intracranial aneurysms (IAs) remains challenging. Therefore, hemodynamic simulations are increasingly applied toward supporting physicians during treatment planning. However, due to several assumptions, the clinical acceptance of these methods remains limited. METHODS: To provide an overview of state-of-the-art blood flow simulation capabilities, the Multiple Aneurysms AnaTomy CHallenge 2018 (MATCH) was conducted. Seventeen research groups from all over the world performed segmentations and hemodynamic simulations to identify the ruptured aneurysm in a patient harboring five IAs. Although simulation setups revealed good similarity, clear differences exist with respect to the analysis of aneurysm shape and blood flow results. Most groups (12/71%) included morphological and hemodynamic parameters in their analysis, with aspect ratio and wall shear stress as the most popular candidates, respectively. RESULTS: The majority of groups (7/41%) selected the largest aneurysm as being the ruptured one. Four (24%) of the participating groups were able to correctly select the ruptured aneurysm, while three groups (18%) ranked the ruptured aneurysm as the second most probable. Successful selections were based on the integration of clinically relevant information such as the aneurysm site, as well as advanced rupture probability models considering multiple parameters. Additionally, flow characteristics such as the quantification of inflow jets and the identification of multiple vortices led to correct predictions. CONCLUSIONS: MATCH compares state-of-the-art image-based blood flow simulation approaches to assess the rupture risk of IAs. Furthermore, this challenge highlights the importance of multivariate analyses by combining clinically relevant metadata with advanced morphological and hemodynamic quantification.

High-Frequency Fluctuations in Post-stenotic Patient Specific Carotid Stenosis Fluid Dynamics: A Computational Fluid Dynamics Strategy Study.

PURPOSE: Screening of asymptomatic carotid stenoses is performed by auscultation of the carotid bruit, but the sensitivity is poor. Instead, it has been suggested to detect carotid bruit as neck's skin vibrations. We here take a first step towards a computational fluid dynamics proof-of-concept study, and investigate the robustness of our numerical approach for capturing high-frequent fluctuations in the post-stenotic flow. The aim was to find an ideal solution strategy from a pragmatic point of view, balancing accuracy with computational cost comparing an under-resolved direct numerical simulation (DNS) approach vs. three common large eddy simulation (LES) models (static/dynamic Smagorinsky and Sigma). METHOD: We found a reference solution by performing a spatial and temporal refinement study of a stenosed carotid bifurcation with constant flow rate. The reference solution [Formula: see text] was compared against LES for both a constant and pulsatile flow. RESULTS: Only the Sigma and Dynamic Smagorinsky models were able to replicate the flow field of the reference solution for a pulsatile simulation, however the computational cost of the Sigma model was lower. However, none of the sub-grid scale models were able to replicate the high-frequent flow in the peak-systolic constant flow rate simulations, which had a higher mean Reynolds number. CONCLUSIONS: The Sigma model was the best combination between accuracy and cost for simulating the pulsatile post-stenotic flow field, whereas for the constant flow rate, the under-resolved DNS approach was better. These results can be used as a reference for future studies investigating high-frequent flow features.

The FDA nozzle benchmark: "In theory there is no difference between theory and practice, but in practice there is".

The utility of flow simulations relies on the robustness of computational fluid dynamics (CFD) solvers and reproducibility of results. The aim of this study was to validate the Oasis CFD solver against in vitro experimental measurements of jet breakdown location from the FDA nozzle benchmark at Reynolds number 3500, which is in the particularly challenging transitional regime. Simulations were performed on meshes consisting of 5, 10, 17, and 28 million (M) tetrahedra, with Δt = 10-5 seconds. The 5M and 10M simulation jets broke down in reasonable agreement with the experiments. However, the 17M and 28M simulation jets broke down further downstream. But which of our simulations are "correct"? From a theoretical point of view, they are all wrong because the jet should not break down in the absence of disturbances. The geometry is axisymmetric with no geometrical features that can generate angular velocities. A stable flow was supported by linear stability analysis. From a physical point of view, a finite amount of "noise" will always be present in experiments, which lowers transition point. To replicate noise numerically, we prescribed minor random angular velocities (approximately 0.31%), much smaller than the reported flow asymmetry (approximately 3%) and model accuracy (approximately 1%), at the inlet of the 17M simulation, which shifted the jet breakdown location closer to the measurements. Hence, the high-resolution simulations and "noise" experiment can potentially explain discrepancies in transition between sometimes "sterile" CFD and inherently noisy "ground truth" experiments. Thus, we have shown that numerical simulations can agree with experiments, but for the wrong reasons.

Real-World Variability in the Prediction of Intracranial Aneurysm Wall Shear Stress: The 2015 International Aneurysm CFD Challenge.

PURPOSE: Image-based computational fluid dynamics (CFD) is widely used to predict intracranial aneurysm wall shear stress (WSS), particularly with the goal of improving rupture risk assessment. Nevertheless, concern has been expressed over the variability of predicted WSS and inconsistent associations with rupture. Previous challenges, and studies from individual groups, have focused on individual aspects of the image-based CFD pipeline. The aim of this Challenge was to quantify the total variability of the whole pipeline. METHODS: 3D rotational angiography image volumes of five middle cerebral artery aneurysms were provided to participants, who were free to choose their segmentation methods, boundary conditions, and CFD solver and settings. Participants were asked to fill out a questionnaire about their solution strategies and experience with aneurysm CFD, and provide surface distributions of WSS magnitude, from which we objectively derived a variety of hemodynamic parameters. RESULTS: A total of 28 datasets were submitted, from 26 teams with varying levels of self-assessed experience. Wide variability of segmentations, CFD model extents, and inflow rates resulted in interquartile ranges of sac average WSS up to 56%, which reduced to < 30% after normalizing by parent artery WSS. Sac-maximum WSS and low shear area were more variable, while rank-ordering of cases by low or high shear showed only modest consensus among teams. Experience was not a significant predictor of variability. CONCLUSIONS: Wide variability exists in the prediction of intracranial aneurysm WSS. While segmentation and CFD solver techniques may be difficult to standardize across groups, our findings suggest that some of the variability in image-based CFD could be reduced by establishing guidelines for model extents, inflow rates, and blood properties, and by encouraging the reporting of normalized hemodynamic parameters.

Multiple Aneurysms AnaTomy CHallenge 2018 (MATCH): Phase I: Segmentation.

PURPOSE: Advanced morphology analysis and image-based hemodynamic simulations are increasingly used to assess the rupture risk of intracranial aneurysms (IAs). However, the accuracy of those results strongly depends on the quality of the vessel wall segmentation. METHODS: To evaluate state-of-the-art segmentation approaches, the Multiple Aneurysms AnaTomy CHallenge (MATCH) was announced. Participants carried out segmentation in three anonymized 3D DSA datasets (left and right anterior, posterior circulation) of a patient harboring five IAs. Qualitative and quantitative inter-group comparisons were carried out with respect to aneurysm volumes and ostia. Further, over- and undersegmentation were evaluated based on highly resolved 2D images. Finally, clinically relevant morphological parameters were calculated. RESULTS: Based on the contributions of 26 participating groups, the findings reveal that no consensus regarding segmentation software or underlying algorithms exists. Qualitative similarity of the aneurysm representations was obtained. However, inter-group differences occurred regarding the luminal surface quality, number of vessel branches considered, aneurysm volumes (up to 20%) and ostium surface areas (up to 30%). Further, a systematic oversegmentation of the 3D surfaces was observed with a difference of approximately 10% to the highly resolved 2D reference image. Particularly, the neck of the ruptured aneurysm was overrepresented by all groups except for one. Finally, morphology parameters (e.g., undulation and non-sphericity) varied up to 25%. CONCLUSIONS: MATCH provides an overview of segmentation methodologies for IAs and highlights the variability of surface reconstruction. Further, the study emphasizes the need for careful processing of initial segmentation results for a realistic assessment of clinically relevant morphological parameters.