A Spatiotemporal exploration and 3D modeling of blood flow in healthy carotid artery bifurcation from two modalities: Ultrasound-Doppler and phase contrast MRI☆
Introduction
Vascular maladies may be caused by thrombi and can lead to stroke [1]. Nowadays, this pathology represents a major health challenge since it is one of the leading causes of mortality all over the world (5.5 million people died of stroke in 2016) [2]. Hence, the importance of investigating the blood flow pattern, notably in the carotid artery which irrigates the brain. This artery is located in the neck. It is composed of the common carotid artery (CCA) which divides into the external carotid artery (ECA) and the internal carotid artery (ICA) irrigating the face and the brain respectively. Several in vivo [3], [4], [5], [6], [7], in vitro [8], [9], [10], in-silico [11], [12], [13], [14], [15], and mixed image-CFD (Computational Fluid Dynamics) [16], [17] approaches have focused on this artery and associated anomalies. The investigations are mainly aimed at localizing and characterizing vessel pathologies from velocity distribution.
CFD simulation has received considerable interest owing to its ability to give access to parameters characterizing the vascular flow, not easily accessible through direct measurements, with notably Wall Shear Stress (WSS) parameters, Oscillatory Shear Index (OSI), Relative Residence Time (RRT) and helicity. Several studies have pointed the importance of having an accurate carotid geometry [18], [19]. Some studies investigated the impact of various rheological models in CFD modeling [14], [20]. Morbiducci and al. [14] suggested that if the rheological model was simplified and blood attributed a constant viscosity (Newtonian fluid) instead of a variable one (Non Newtonian), the difference in the simulations would be less than 10%. Other works [18], [21] have shown that inlet boundary conditions significantly affect the numerical simulation of velocity, which also depends on the carotid artery localization.
Blood flow analysis can confirm the presence of a local vessel anomaly, its impact on the blood flow pattern and its possible evolution [18], [22], [23], [24]. Some flow indicators such as TAWSS (time averaged wall shear stress) and OSI can help detect and localize abnormal WSS increase leading to thrombosis formation and stroke [24].
Realistic image-based patient specific CFD modeling requires the extraction of several pieces of information from medical data with at least: (i) the vascular morphology from medical imaging (Computed tomography Angiography (CTA) [25] or Magnetic Resonance Imaging (MRI) [26]). (ii) measurements of blood flow velocity (Ultrasound Doppler (US-Doppler) [27] or phase contrast Magnetic Resonance (PC-MRI) [26]) to provide at least arterial input functions.
Regarding the anatomy extraction, clinicians generally use CTA since it offers a higher spatial resolution than MRI and has proved its reliability [28], especially when determining a stenosis degree. As for velocity, the most widespread clinical assessment is based on US-Doppler because of its accessibility and its ease of use, particularly for the neck. However, this examination is operator-dependent and getting velocity values all over the carotid artery bifurcation is a hard task. This limitation may be overcome with PC-MRI which can bring both geometric and hemodynamic information simultaneously thanks to a compromise between spatial resolution and acquisition time. In this context, few studies compared velocity measurements in the carotid artery between US-Doppler and PC-MRI [3], [4], [5], [29]. They showed that there was a significant variation of velocity values in the CCA and that PC-MRI generally leads to smaller velocity values compared with US-Doppler. However, these studies have not considered CFD simulations to complete measurement characteristics in order to obtain an overall carotid hemodynamic exploration.
In this paper, our objective is to design an optimized patient specific CFD workflow in terms of quality of velocity profiles, computational efficiency and clinical applicability. The anatomical data are obtained from PC-MRI, while velocity data are obtained from both PC-MRI and US-Doppler. This allows us to address two sub-goals (i) comparing velocity measurements from the two imaging modalities implying different acquisition conditions (the widespread US-Doppler, and less routine PC-MRI) and providing different velocity quantifications (1D-axial vs 3D). (ii) investigating the optimal combination of velocity measurements and CFD modeling to provide realistic patient specific flow simulations. We investigated and modeled blood flow velocity from imaging data in ten healthy volunteers. This work may also allow deriving PC-MRI and/or US-Doppler measured velocities given numerical velocities. Section 2 describes the imaging data sets and the analysis methodology. Section 3 presents the obtained results that are discussed in Section 4.
Section snippets
Material and methods
The arterial geometrical model was extracted from MR anatomical images and the hemodynamic modeling was performed from the obtained models. The velocity waveforms from both PC-MRI and US-Doppler were extracted at three locations: the right CCA, ECA and ICA.
Results
Velocity waveforms were extracted from PC-MRI ( and ) and Doppler-US () imaging data for ten volunteers at the three localizations of the carotid artery. Two velocity profiles were simulated. The first one, called , is based on PC-MRI data profile used as inlet boundary condition. The second one, termed , results from a simulation with profile as inlet boundary condition. All velocity waveforms characteristics are reported in the Appendix (
Discussion
In this study, the PC-MRI velocity behavior in the carotid bifurcation was investigated in time and space. It was observed that the velocity estimation increased with the window size up to maximum (46 × 3, Fig. 3). We calculated the global velocity difference between located in the three key sites mentioned before (Fig. 3) and according to each window size (Fig. 12). We found that for the three arteries, the velocity difference presented a piecewise linear growth before
Conclusion
In this paper, we investigated blood velocity quantification over the carotid artery bifurcation from PC-MRI and US-Doppler, and 3D velocity modeling from these modalities. We performed a hemodynamic modeling on 10 healthy subjects and assessed the impact of using each of these imaging modalities as inlet boundary condition onto the simulated velocities over the carotid artery. We observed that the US-Doppler generally leads to higher velocity values compared to PC-MRI, which has also been
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Arij Debbich completed her engineering’s degree in real time computer science from ISSAT SousseTunisia and her Master’s degree in computer science and image processing from ENSI-Tunisia. She is currently a Ph.D. student at ENIS, University of Sfax and Laboratory of Technology and Medical Imaging LTIM, Faculty of Medicine, University of Monastir. Her current research interests include blood flow modeling and image analysis.
References (55)
- et al.
Accuracy and precision of a plane wave vector flow imaging method in the healthy carotid artery
Ultrasound Med. Biol.
(2018) - et al.
Common carotid artery flow measured by 3-D ultrasonic vector flow imaging and validated with magnetic resonance imaging
Ultrasound Med. Biol.
(2017) - et al.
Carotid artery flow as determined by real-time phase-contrast flow MRI and neurovascular ultrasound: A comparative study of healthy subjects
Eur. J. Radiol.
(2018) - et al.
Hemodynamic effects of long-term morphological changes in the human carotid sinus
J. Biomech.
(2015) - et al.
Transitional flow analysis in the carotid artery bifurcation by proper orthogonal decomposition and particle image velocimetry
Med. Eng. Phys.
(2013) - et al.
Pulsatile poststenotic flow studies with laser Doppler anemometry
J. Biomech.
(1984) - et al.
Noninvasive assessment of carotid artery stenoses by the principle of multiscale modelling of non-Newtonian blood flow in patient-specific models
Appl. Math. Comput.
(2018) - et al.
A comparison of 4D flow MRI-derived wall shear stress with computational fluid dynamics methods for intracranial aneurysms and carotid bifurcations—a review
Magn. Reson. Imaging
(2018) - et al.
On the importance of blood rheology for bulk flow in hemodynamic models of the carotid bifurcation
J. Biomech.
(2011) - et al.
Analysis of the effects of different pulsatile inlet profiles on the hemodynamical properties of blood flow in patient specific carotid artery with stenosis
Comput. Biol. Med.
(2013)
Image-based computational hemodynamics evaluation of atherosclerotic carotid bifurcation models
Comput. Biol. Med.
Computational fluid dynamics to evaluate the management of a giant internal carotid artery aneurysm
World Neurosurg.
2091330 volumetric carotid blood flow assessment-a comparison of US and MRI
Ultrasound Med. Biol.
Delaunay triangulation in computational fluid dynamics
Comput. Math. Appl.
Validation of a 1D patient-specific model of the arterial hemodynamics in bypassed lower-limbs: simulations against in vivo measurements
Med. Eng. Phys.
Pulse wave propagation in a model human arterial network: assessment of 1-D visco-elastic simulations against in vitro measurements
J. Biomech.
Blood flow modeling in carotid arteries with computational fluid dynamics and MR imaging
Academic Radiol.
Numerical flow studies in human carotid artery bifurcations: basic discussion of the geometric factor in atherogenesis
J. Biomed. Eng.
The influence of the non-Newtonian properties of blood on the flow in large arteries: steady flow in a carotid bifurcation model
J. Biomech.
Optical coherence tomography-based patient-specific coronary artery reconstruction and fluid–structure interaction simulation
Biomech. Model. Mechanobiol.
Cause of death
Our World Data
Comparison of blood flow velocity quantification by 4D flow MR imaging with ultrasound at the carotid bifurcation
Am. J. Neuroradiol.
Characterization of common carotid artery blood-flow waveforms in normal human subjects
Physiol. Meas.
4D flow MRI
J. Magn. Reson. Imaging
Modelling and simulation of low-density lipoprotein transport through multi-layered wall of an anatomically realistic carotid artery bifurcation
J. R. Soc. Interface
Hemodynamics analysis of patient-specific carotid bifurcation: A CFD model of downstream peripheral vascular impedance
Int. J. Numer. Methods Biomed. Eng.
On recent progress in modelling and simulations of multi-scale transfer of mass, momentum and particles in bio-medical applications
Flow Turbul. Combust.
Cited by (0)
Arij Debbich completed her engineering’s degree in real time computer science from ISSAT SousseTunisia and her Master’s degree in computer science and image processing from ENSI-Tunisia. She is currently a Ph.D. student at ENIS, University of Sfax and Laboratory of Technology and Medical Imaging LTIM, Faculty of Medicine, University of Monastir. Her current research interests include blood flow modeling and image analysis.
Asma Ben Abdallah was born in Tunisia on March 20, 1972. She hold her Ph.D. in computer science from the National Institute of Computer Science (ENSI) in Tunisia in 2007. Since 2004, she was an Assistant Professor on informatics at the High Institute of Application Science and Technology in Sousse. Since 2008, she was an Assistant Professor on informatics at the Higher Institute of Informatics and Mathematics of Monastir (ISIMM), in Tunisia and Since 2018, she is Professor on informatics at the same institution (ISIMM). On research, she works in the Laboratory of Technology and Medical Imaging (LTIM), at the Faculty of Medicine of Monastir, University of Monastir, Tunisia. Her current interests include computer vision-imaging.
Mezri Maatouk was born in 1979 in Monastir, Tunisia. He is a medical doctor (FMM-University of Monastir in 2007). He obtained his M.Res. Signals and Images in Medicine at the University of Créteil, France, in 2009. He is Associate Professor in radiology at the Faculty of Medicine of Monastir.
Badii Hmida was born in 1979 in Mahdia, Tunisia. He is a medical doctor (FMM-University of Monastir in 2009); He is Associate Professor in radiology at the Faculty of Medicine of Monastir. He obtained a University Diploma of cardiovascular imaging (University of Bichat in 2017).
Monica Sigovan, received a Ph.D. in Biomedical Engineering in 2009 from the University Claude Bernard Lyon 1. Since 2016, she is a CNRS Research Scientist and part of the CREATIS laboratory, University of Lyon, CNRS UMR 5220, Inserm U1206, Lyon, France. The primary focus of her work is MR image acquisition and postprocessing, development of quantitative imaging markers for vascular disease and its progression.
Patrick Clarysse received a MSc degree in 1987 and a Ph.D. in 1991 from the Scientific and Technological University of Lille, France. Since 1992, he is with the French National Center for Scientific Research (CNRS) at CREATIS, University of Lyon, CNRS UMR 5220, Inserm U1206, Lyon, France. Currently, he is Research Director and heads the ‘Modeling and Imaging of Vessels, Thorax and Brain’ group. His primary research interests are in the bioengineering and medical image analysis fields, and include medical image processing workflows, multidimensional/multimodal image segmentation and registration, motion estimation and deformable models with applications to the 3D analysis and modeling of heart functions and motion of thoracic structures. Patrick Clarysse has coauthored more than 140 papers in international journals and conferences in the field and 3 books. He is a member of IEEE, EMBS and of the French Society of BioEngineering (SFGBM).
Mohamed Hédi Bedoui was born in 1965 in Teboulba, Tunisia. He is an engineer in electronics (ENIS-University of Sfax in 1988). He obtained his Ph.D. Biomedical Engineer at the University of Lille, France, in 1993. He is Professor in Biophysics at the Faculty of Medicine of Monastir, Director of the laboratory LTIM (Technology and Medical Imaging)-LR12ES06 and President of the Society Tunisian Association for the Promotion of Applied Research-ATUPRA. His current interests include modeling and processing medical data from the sensor to optimizing the implementation of embedded systems.
- ☆
This study was conducted within the framework of the LABEX PRIMES (ANR-11-LABX-0063) project of the University of Lyon, within the “Investissements d’Avenir”(ANR-11-IDEX-0 0 07) program operated by the French National Research Agency (ANR) .