- Open Access
Numerical prediction of thrombus risk in an anatomically dilated left ventricle: the effect of inflow cannula designs
© The Author(s) 2016
- Published: 28 December 2016
Implantation of a rotary blood pump (RBP) can cause non-physiological flow fields in the left ventricle (LV) which may trigger thrombosis. Different inflow cannula geometry can affect LV flow fields. The aim of this study was to determine the effect of inflow cannula geometry on intraventricular flow under full LV support in a patient specific model.
Computed tomography angiography imaging of the LV was performed on a RBP candidate to develop a patient-specific model. Five inflow cannulae were evaluated, which were modelled on those used clinically or under development. The inflow cannulae are described as a crown like tip, thin walled tubular tip, large filleted tip, trumpet like tip and an inferiorly flared cannula. Placement of the inflow cannula was at the LV apex with the central axis intersecting the centre of the mitral valve. Full support was simulated by prescribing 5 l/min across the mitral valve. Thrombus risk was evaluated by identifying regions of stagnation. Rate of LV washout was assessed using a volume of fluid model. Relative haemolysis index and blood residence time was calculated using an Eulerian approach.
The inferiorly flared inflow cannula had the lowest thrombus risk due to low stagnation volumes. All cannulae had similar rates of LV washout and blood residence time. The crown like tip and thin walled tubular tip resulted in relatively higher blood damage indices within the LV.
Changes in intraventricular flow due to variances in cannula geometry resulted in different stagnation volumes. Cannula geometry does not appreciably affect LV washout rates and blood residence time. The patient specific, full support computational fluid dynamic model provided a repeatable platform to investigate the effects of inflow cannula geometry on intraventricular flow.
- Heart failure
- Left ventricular assist device
- Multiphase modelling
The number of patients receiving left ventricular assist devices (LVADs) to treat heart failure is on the rise, with over 2500 devices implanted per year in the US alone . Inflow attachment of LVADs is generally achieved via the left ventricular apex with a special tube called an inflow cannula (IC), while outflow attachment is generally achieved with a graft attached to the ascending aorta. Insertion of a foreign object into the left ventricle (LV) can affect intraventricular flow. This disruption to ventricular haemodynamics can potentially cause thrombus formation due to stagnation zones, recirculation zones and high shear stresses [2–7]. Currently, no gold standard exists for the IC geometry, with each clinically used device having a custom design.
To minimise thrombus formation, it is beneficial to have low LV blood residence time (BRT), low stagnation and efficient LV washout. Blood stagnation and recirculation can be related to low velocities and shear rates . Studies have shown that the native haemodynamics inside the LV are asymmetric, unsteady, and have a large diastolic vortex that directs blood towards the aortic valve [8–11]. However, most computational fluid dynamic (CFD) studies of the LV have been extensively simplified [12–16]. Most models do not consider recreating natural ventricular geometry, haemodynamics and the corresponding effects of an IC. This could significantly alter the areas where thrombus formation may occur.
Inflow cannula geometry, insertion depth, insertion angle and position may all alter the intraventricular flow fields. One study has shown that by increasing the IC insertion length from 24 to 34 mm, a higher survival rate of 63.5% compared to 52.9% was observed . This minor difference in insertion length can greatly affect the patient´s outcome; however, the exact reason is unknown. The incidence of stroke was found to be 23.2 and 3.8% with the short and long cannula, respectively . This significant drop in neurological adverse events could stem from a range of contributing factors including patient specific causes, surgical technique and blood flow patterns. It was hypothesised that the difference in neurological events could be linked to poor LV washout and stagnation regions caused by the shorter IC.
While there are several studies reporting on intraventricular flow dynamics with LVAD support, these are limited by the lack of anatomically correct ventricular geometry. For instance, Loerakker et al.  and Ong et al.  simulated the LV geometry using a prolate ellipsoid with tubes either side to represent the atrium, aorta and LVAD IC. Ong et al.  studied the effect of cannula placement on thrombosis. Three different insertion lengths of a trumpet tipped inflow cannula were simulated where the tip was near the apex, inserted one-fourth of the LV length, or inserted half of the LV length. The risk of thrombosis was evaluated by intraventricular vorticity distributions, intensities, stagnations and wall shear stresses. It was found that the cannula inserted one-fourth into the LV resulted in negligible fluid stagnation as it created higher vortex intensities. Even though a deformable wall was implemented, the non-physiological intraventricular flows provided by the inlet boundary condition and ventricular geometry could affect the results significantly.
Liu et al.  used a conical volume with no valves to represent the LV to investigate 4 different types of cannulae: blunt, bevelled, trumpet and caged. Blood compatibility was evaluated based on wall shear stress and exposure time, of which none met the estimated criteria for haemolysis. It was found that the trumpet design had the best blood compatibility, comparatively.
An analysis of thrombosis potential in LVAD drainage cannulae was investigated by Fraser et al. . The study implemented three clinically available, at the time of writing, cannulae (Medtronic DLP 12, 16 and 24 F) in an MRI segmented LV model. Different flow rates were used, applied at a constant uniform velocity. It was found that 12 and 16 F cannulae were superior due to lower fluid stagnation volumes compared to the 24 F cannula at flows below 0.75 l/min.
The impact of mitral valve (MV) modelling should also not be underestimated, as this dictates the inlet flow to the ventricle. Others have either ignored the MV entirely [12–14] or modelled it using rigid plates . Most limitations in previous studies arose from numerical models that could not recreate physiological LV haemodynamics, generally attributed to over-simplification. Therefore, a model which combines anatomically correct ventricular and mitral valve geometry would provide beneficial evaluation of LVAD IC geometry. The aim of this study was to create a patient-specific LV with an approximated MV model to determine the effects of various IC geometries on intraventricular flow in a total heart failure simulation. It was hypothesised that a cannula with a smoother transition from the cannula to the endocardium can reduce regions of stagnant flow and increase LV washout.
The left atrium was represented by a 40 mm diameter cylinder with the MV placed adjacent to the aortic valve with guidance from CT data. The aortic valve was not included in the model due to full support by the LVAD: the aortic valve is always closed.
Mitral valve modelling
Mesh statistics, as reported by ANSYS fluent
Number of cells
Minimum orthogonal quality
Blood was defined to be Newtonian fluid with a viscosity and density of 0.0035 Pa s and 1059 kg/m3, respectively. A constant flow rate of 5 l/min was prescribed at the inlet (atrium). Constant flow was chosen to represent full LVAD support with no native ejection through the aortic valve. It was assumed that whichever rotary LVAD is selected, the rotational speed would be set to achieve a flow rate of 5 l/min. This flow rate was equivalent to a bulk mass flow rate of 0.088 kg/s with the flow prescribed normal to the boundary. The cannula outlet had a 0 Pa pressure outlet. All walls were defined with a no slip condition.
From the low Re number, a laminar model was used. ANSYS Fluent was used for the calculations. The Pressure Implicit with Splitting of Operator (PISO) pressure–velocity coupling scheme was selected. Spatial discretisation for pressure, momentum and volume fraction was defined to be a second order upwind scheme. A transient simulation of 20 s was performed. Time integration was done using a bounded second order implicit backward Euler method with time steps of 0.002 s. Results for analysis were taken from the last 15 s. The first 5 s allowed the flow field to be established. Results were deemed to be converged when the scaled residuals were below 10−4 for continuity, x velocity, y velocity and z velocity. Each time step required approximately 15–20 iterations. Using 16 CPUs per model, each simulation took approximately 1.5 weeks to complete. All calculations were performed on a High Performance Computing cluster (Queensland University of Technology, Brisbane, Australia).
There is currently no standard in the definition of stagnation volume [13, 14, 20, 21]. Regions of potential thrombus formation were defined by volumes of low velocity magnitudes, less than 0.001 m/s, and strain rates of less than 2 s−1 . The additional low strain rate criteria prevented the inclusion of low velocity magnitudes experienced at the walls. This criteria was used to relatively compare the models. All cell volumes that met the criteria were assigned a value of 1, all others were assigned 0. A plane was defined 50 mm above the apex and all volumes below this plane were defined as a volume of interest. This was used for calculation of thrombus risk (TR) volumes around the cannula. The TR volume was calculated by integrating the assigned cell value, 0 or 1, over the volume of interest. Evaluation of the stagnation regions was performed at time points 0, 5, 10 and 15 s. At each time point, an image was captured that resulted in four images per cannula. Using imaging processing, the median channel values for each pixel of the four images were taken, resulting in removal of stagnation regions that were intermittent. This allowed clearer qualitative comparisons between the cannulae.
Volume of fluid modelling
Volume of fluid (VOF) modelling, as developed by Sonntag et al. , was used to determine LV washout. The VOF model is generally used for interface tracking of immiscible fluids. However, in this study the VOF model was used to calculate displacement of one fluid (old blood) by another fluid (new blood), and so the primary and secondary phase had identical fluid properties. An implicit formulation was applied with dispersed interface modelling. The secondary phase was started after 5 s of flow stabilisation. The pressure outlet had a secondary phase backflow volume fraction of 0, which prevents “new blood” from entering from the outlet. The time taken for LV washout was calculated by taking the volume integral of the volume fraction of each cell, which ranged between 0 and 1. If the initial phase (old blood) was of interest, then a volume fraction of 1 indicates the cell is full of old blood and conversely, 0 indicates a cell has no old blood. A value of 0.5 indicates the surface between the new and old blood. The rate of LV washout was determined by integrating the volume fraction, between 0 and 1, over the cell volume. This was normalised with the total volume of the entire domain.
Mitral jet movement
Mean angle between jet and cannula (°)
Standard deviation (°)
Volume of fluid washout model
Haemolysis index and blood residence time
Cannula wall shear stress
Average wall shear stress on cannula surface
Mean WSS (Pa)
Standard deviation (Pa)
This study aimed to determine the effects of different IC geometries on left intraventricular flow in terms of TR, BRT and LV washout. TR was defined by low velocity magnitudes coupled with low strain rate. In this study, it was shown that the inferiorly flared cannula has the potential to reduce incidences of thrombus formation with comparatively lower stagnation regions while all cannulae resulted in similar LV washout rates and blood residence times.
The implementation of a static approximation of a MV, developed by Domenichini et al. , was able to recreate general physiological flows. Intraventricular flow profiles were similar to a study by Bermejo et al. , where echocardiographic data was obtained in nonischemic dilated cardiomyopathy patients. A large dominant clockwise rotation was seen with all cannula cases, which agreed with the echocardiographic data during the diastasis phase . Moreover, the low velocity region under the posterior MV leaflet agreed with the study by Bermejo et al. . When using phase contrast magnetic imaging, it has been shown that a counter clockwise rotation forms during diastole near the apex . Since the apex was cannulated in this study, a similar counter clockwise rotation of blood was seen superior to the LV apex and closer to the tip of the MV jet.
It has been suggested that the mid-diastolic vortices enhance blood mixing which can prevent stagnation areas . Qualitatively evaluating the streamlines in Fig. 5, it can be seen that for cannula E, there were more distinct vortices near the ventricular apex which support the suggestion of enhanced blood mixing to reduce stagnation areas. Since this study modelled total heart failure, there were no distinct phases of the cardiac cycle. In the present work, all cannulae resulted in scattered vortex cores, which agrees with previous studies that have reported disturbed vortex structures in dilated ventricles [30–33].
The highest and most consistent areas of TR volume occurred at the interface between the cannula and endocardium where wedge thrombus has been reported in clinic . Thrombus formation can be caused by hyper coagulability, stasis and endothelial injury, as described by Virchow’s triad . As endothelial injury is already sustained around the ventricular apex, there is a need to reduce stasis or coagulability. Cannula E had the most potential to significantly reduce incidences of wedge thrombus formation by minimising areas of stasis, while its flared design may also cover the area of endothelial injury.
When comparing cannulae A, B and C to cannula E; the shorter cannula appeared to reduce stagnation areas as similarly reported by Ong et al.  and Antaki et al. . However, cannula D had a similar length to cannula B and was found to have relatively low stagnation regions. This may indicate that cannula length is not the only dependent variable on stagnation regions: cannula geometry can have an effect.
The use of volume of fluid model was an efficient way to calculate LV washout. It was found that all cannulae had similar LV washout rates, indicating that different IC geometries do not significantly affect LV washout. It was expected that cannulae with higher stagnation regions would result in slower LV washout, however this was not found. This could be attributed to the low stagnation volumes and hence the volume of fluid may not have the sensitivity to detect such low volume differences. Furthermore, the blood residence time, using the Eulerian method, was found to be similar between the five cannulae, which supports the findings with similar rates of LV washout.
The highest time-averaged WSS was found with cannula A at 2.1 Pa, which is approximately 100 orders of magnitude lower than the suspected haemolysis threshold found by Reinhard et al. . Between the highest (cannula A) and lowest (cannula B) time-averaged WSS, the difference was approximately 47%. The major difference between cannula A and B was that cannula A has a large number of surface features which are facing superiorly, representing a crown shape, while cannula B has a relatively sharp and simple inflow edge.
In general, the difference in HI due to different IC tips indicated that the complex crown shape (cannula A) and sharp inlet (cannula B) promotes relatively higher shear stresses within the LV as it has been found that the blood residence time was similar across all cannulae. It must be noted that the implemented HI method has been shown to have good correlation with experimental data but has large errors . As a consequence, the HI prediction was used to relatively compare the models.
In summary, all cannulae had similar rates of LV washout and supported by similar blood residence times. The risk of thrombus formation due to blood stagnation is not solely dependent on cannula insertion length but also on cannula geometry. Cannula E was found to have the lowest risk of thrombus formation which can be attributed to the smoother transition from endocardium to the IC surface. As a result, it has been shown that the inferiorly flared cannula E has the potential to reduce the likelihood for wedge thrombus formation.
One major limitation of the presented model was the static LV and constant inflow. In most instances, some contractility of the LV is still present. The addition of a deforming wall boundary would change regions of possible stagnation and rate of LV washout. Meanwhile, implantation of LVADs are often performed with a functioning right ventricle. As a result, the incoming flow from the pulmonary vein would still have some pulsatility. This acceleration and deceleration of blood may affect the TR regions, LV washout and WSS on the cannula. Furthermore, the outlet boundary condition would not be 0 Pa in a pulsatile system, and a method to better represent physiological boundaries would be the coupling of a lumped parameter model, which numerically describes the cardiovascular system . Clinically, the LVAD is often set to a speed where some aortic valve opening occurs to prevent valve fusion, aortic incompetence and reduction in thrombus potential . As a result, an additional outlet boundary at the aorta could affect the intraventricular flow. Furthermore, a static MV was implemented. Due to adaptive ventricular remodelling from cardiac insufficiency, the geometry of the MV will change [40–42]. These changes could alter the intraventricular flow. The inclusion of realistic heart wall and valve movements is highly complex but can be achieved through fluid structure interactions or a prescribed moving mesh. The main limitation of such models is the excessive computational time required. Using such models to compare a large variety of cannulae would be computationally unfeasible at this time. This study did not take into account different apical thicknesses, which would affect insertion length. Future evaluations should consider insertion length and position, not just geometry. The presented CFD model did not include internal LV features but has been shown to increase the size of recirculation zones and deeper MV jet penetration .
Clinical translation of these results must first consider the given limitations. In the case of total heart failure with a severely dilated LV, the results from this study could provide a general overview of the intraventricular flow. This study can aid in the future design of inflow cannulae. It was found to be beneficial to minimise the inlet tip complexity, which can improve haemolysis indices within the LV. To minimise the risk of thrombus due to stagnation, an increase in angle between the endocardium and inflow cannula wall could be advantageous. Further validation of these recommendations will be required by using other techniques such as particle image velocimetry and in vivo validation.
Five different IC were evaluated in a patient specific LV to determine blood stagnation regions, LV washout, relative blood damage indices, BRT, and average cannula WSS. General intraventricular flow features were able to be reproduced with an approximated MV. Areas of blood stagnation (TR volumes) were consistently found at the interface between the IC and endocardium. All cannulae had similar rates of LV washout and RBT. The complex crown tip results in relatively higher blood damage indices. However, all cannula WSS were below the haemolysis threshold. The use of an inferiorly flared IC has the potential to lower the incidences of thrombus formation.
SL was responsible for generation the models, performing simulations and writing of the manuscript. BS and MN was responsible for aiding in developing the computational model, data analysis and data interpretation. TK contributed by aiding the design of methods for data analysis. ZL, MW and SG was responsible for the conception and design of the study. All authors have made substantial contribution to drafting and critically reviewing the manuscript. All authors read and approved the final manuscript.
The authors would like to recognize the financial assistance provided by the Australian Postgraduate awards, The Prince Charles Hospital Foundation (NI2014-127, PRO2014-08), the National Health and Medical Research Council Centre for Research Excellence (APP1079421) and the Australian Research Council (LP130100461). Computational resources and services used in this work were provided by the HPC and Research Support Group, Queensland University of Technology, Brisbane, Australia.
The authors declare that they have no competing interests.
About this supplement
This article has been published as part of BioMedical Engineering OnLine Volume 15 Supplement 2, 2016. Computational and experimental methods for biological research: cardiovascular diseases and beyond. The full contents of the supplement are available online http://biomedical-engineering-online.biomedcentral.com/articles/supplements/volume-15-supplement-2.
Availability of data and materials
Data available upon request.
Publication of this article was funded by Queensland University of Technology.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Kirklin JK, Naftel DC, Pagani FD, Kormos RL, Stevenson LW, Blume ED, et al. Seventh INTERMACS annual report: 15,000 patients and counting. J Hear Lung Transplant. 2015;34:1495–504.View ArticleGoogle Scholar
- Bluestein D. Research approaches for studying flow-induced thromboembolic complications in blood recirculating devices. Expert Rev Med Devices. 2004;1:65–80.View ArticleGoogle Scholar
- Fukuda N, Itatani K, Kimura K, Ebihara A, Negishi K, Uno K, et al. Prolonged vortex formation during the ejection period in the left ventricle with low ejection fraction: A study by vector flow mapping. J Med Ultrason. 2014;41:301–10.View ArticleGoogle Scholar
- Folie BJ, Mclntire LV. Mathematical analysis of mural thrombogenesis. Biophys J. 1989;56:1122–41.View ArticleGoogle Scholar
- Petschek H, Adamis D, Kantrowitz AR. Stagnation flow thrombus formation. Trans Am Soc Artif Intern Organs. 1968;14:256–60.Google Scholar
- Alevriadou BR, Moake JL, Turner NA, Ruggeri ZM, Folie BJ, Phillips MD, et al. Real-time analysis of shear-dependent thrombus formation and its blockade by inhibitors of von Willebrand factor binding to platelets. Blood Am Soc Hematol. 1993;81:1263–76.Google Scholar
- Slack SM, Turitto VT. Fluid dynamic and hemorheologic considerations. Cardiovasc. Pathol. 1993;2:11–21.View ArticleGoogle Scholar
- Kheradvar A, Houle H, Pedrizzetti G, Tonti G, Belcik T, Ashraf M, et al. Echocardiographic particle image velocimetry: A novel Technique for quantification of left ventricular blood vorticity pattern. J Am Soc Echocardiogr. 2010;23:86–94.View ArticleGoogle Scholar
- Pedrizzetti G, Domenichini F. Nature optimizes the swirling flow in the human left ventricle. Phys Rev Lett. 2005;95:1–4.View ArticleGoogle Scholar
- Kim WY, Walker PG, Pedersen EM, Poulsen JK, Oyre S, Houlind K, et al. Left ventricular blood flow patterns in normal subjects: a quantitative analysis by three-dimensional magnetic resonance velocity mapping. J Am Coll Cardiol. 1995;26:224–38.View ArticleGoogle Scholar
- Rodevand O, Bjornerheim R, Edvardsen T, Smiseth OA, Ihlen H. Diastolic flow pattern in the normal left ventricle. J Am Soc Echocardiogr. 1999;12:500–7.View ArticleGoogle Scholar
- Loerakker S, Cox LGE, van Heijst GJF, de Mol BA, van de Vosse FN. Influence of dilated cardiomyopathy and a left ventricular assist device on vortex dynamics in the left ventricle. Comput Methods Biomech Biomed Eng. 2008;11(6):49–60.View ArticleGoogle Scholar
- Ong C, Dokos S, Chan B, Lim E, Al Abed A, Bin Abu Osman NA, et al. Numerical investigation of the effect of cannula placement on thrombosis. Theor Biol Med Model. 2013;10:35.View ArticleGoogle Scholar
- Liu GM, Chen HB, Luo FL, Zhang Y, Sun HS, Zhou JY et al. Numerical simulation of LVAD inflow cannulas with different tip. Int J Chem Eng. 2012.Google Scholar
- Suuu B, Zhong L, Wang XK, Zhang JM, Tan RS, Allen JC, et al. Numerical simulation of patient-specific left ventricular model with both mitral and aortic valves by FSI approach. Comput Methods Programs Biomed. 2014;113:474–82.View ArticleGoogle Scholar
- Tsukiya T, Toda K, Sumikura H, Takewa Y, Watanabe F, Taenaka Y, et al. Computational fluid dynamic analysis of the flow field in the newly developed inflow cannula for a bridge-to-decision mechanical circulatory support. J Artif Organs. 2011;14:381–4.View ArticleGoogle Scholar
- Schmid C, Jurmann M, Birnbaum D, Colombo T, Falk V, Feltrin G, et al. Influence of inflow cannula length in axial-flow pumps on neurologic adverse event rate: results from a multi-center analysis. J Heart Lung Transplant. 2008;27:253–60.View ArticleGoogle Scholar
- Fraser KH, Zhang T, Taskin ME, Griffith BP, Wu ZJ. Computational fluid dynamics analysis of thrombosis potential in left ventricular assist device drainage cannulae. ASAIO J. 2010;56:157–63.View ArticleGoogle Scholar
- Domenichini F, Pedrizzetti G. Asymptotic model of fluid-tissue interaction for mitral valve dynamics. Cardiovasc Eng Technol. 2014;6:95–104.View ArticleGoogle Scholar
- Wong K, Samaroo G, Ling I, Dembitsky W, Adamson R, Del Álamo JC, et al. Intraventricular flow patterns and stasis in the LVAD-assisted heart. J Biomech. 2014;47:1485–94.View ArticleGoogle Scholar
- Rossini L, Martinez-Legazpi P, Vu V, Fernández-Friera L, Pérez del Villar C, Rodríguez-López S et al. A clinical method for mapping and quantifying blood stasis in the left ventricle. J Biomech. 2015.Google Scholar
- Sonntag SJ, Kaufmann TAS, Büsen MR, Laumen M, Gräf F, Linde T, et al. Numerical washout study of a pulsatile total artificial heart. Int J Artif Organs. 2014;37:241–52.View ArticleGoogle Scholar
- Fraser KH, Zhang T, Taskin ME, Griffith BP, Wu ZJ. A quantitative comparison of mechanical blood damage parameters in rotary ventricular assist devices: shear stress, exposure time and hemolysis index. J Biomech Eng. 2012;134:081002.View ArticleGoogle Scholar
- Taskin ME, Fraser KH, Zhang T, Wu C, Griffith BP, Wu ZJ. Evaluation of eulerian and lagrangian models for hemolysis estimation. ASAIO J. 2012;58:363–72.View ArticleGoogle Scholar
- Giersiepen M, Wurzinger LJ, Opitz R, Reul H. Estimation of shear stress-related blood damage in heart valve prostheses–in vitro comparison of 25 aortic valves. Int J Artif Organs. 1990;13:300–6.Google Scholar
- Slotosch A, Schlanderer J, Kaiser F, Kriegseis J. Heart flow vortices. An application of flow characterization techniques. Fachtagung “Lasermethoden der Strömungsmesstechnik” Karlsruhe. 2014. p. 45/1–45/8.Google Scholar
- Martin MJ, Chung EML, Ramnarine KV, Goodall AH, Naylor AR, Evans DH. Thrombus size and Doppler embolic signal intensity. Cerebrovasc Dis. 2009;28:397–405.View ArticleGoogle Scholar
- Bermejo J, Benito Y, Alhama M, Yotti R, Martínez-Legazpi P, Del Villar CP, et al. Intraventricular vortex properties in nonischemic dilated cardiomyopathy. Am J Physiol Heart Circ Physiol. 2014;306:H718–29.View ArticleGoogle Scholar
- Seo JH, Vedula V, Abraham T, Lardo AC, Dawoud F, Luo H, et al. Effect of the mitral valve on diastolic flow patterns. Phys Fluids. 2014;26:121901.View ArticleGoogle Scholar
- Pedrizzetti G, La Canna G, Alfieri O, Tonti G. The vortex–an early predictor of cardiovascular outcome? Nat Rev Cardiol. 2014;11(9):545–53.View ArticleGoogle Scholar
- Mangual JO, Kraigher-Krainer E, De Luca A, Toncelli L, Shah A, Solomon S, et al. Comparative numerical study on left ventricular fluid dynamics after dilated cardiomyopathy. J Biomech. 2013;46:1611–7.View ArticleGoogle Scholar
- Chan BT, Ong CW, Lim E, Abu Osman NA, Al Abed A, Lovell NH et al. Simulation of left ventricle flow dynamics with dilated cardiomyopathy during the filling phase. Proceedings annual international conference of the IEEE engineering in medicine and biology society (EMBC). 2012. pp. 6289–92.Google Scholar
- Baccani B, Domenichini F, Pedrizzetti G, Tonti G. Fluid dynamics of the left ventricular filling in dilated cardiomyopathy. J Biomech. 2002;35:665–71.View ArticleGoogle Scholar
- Kurihara C, Ono M, Nishimura T, Saito A, Taketani T, Hisagi M, et al. Use of DuraHeart® support for more than 1 year as the first successful bridge to heart transplantation in Japan. J Artif Organs. 2011;14:67–9.View ArticleGoogle Scholar
- Bagot CN, Arya R. Virchow and his triad: a question of attribution. Br J Haematol. 2008;143:180–90.View ArticleGoogle Scholar
- Antaki JF, Dennis TJ, Konishi H, Brown ME, Maher TR, Tomczak JP, et al. An improved left ventricular cannula for chronic dynamic blood pump support. Artif Organs. 1995;19:671–5.View ArticleGoogle Scholar
- Paul R, Apel J, Klaus S, Schuegner F, Schwindke P, Reul H. Shear stress related blood damage in laminar Couette flow. Artif Organs. 2003;27:517–29.View ArticleGoogle Scholar
- Neidlin M, Corsini C, Sonntag SJ, Schulte-Eistrup S, Schmitz-Rode T, Steinseifer U et al. Hemodynamic analysis of outflow grafting positions of a ventricular assist device using closed-loop multiscale CFD simulations: Preliminary results. J Biomech. 2016.Google Scholar
- Tolpen S, Janmaat J, Reider C, Kallel F, Farrar D, May-Newman K. Programmed speed reduction enables aortic valve opening and increased pulsatility in the LVAD-assisted heart. ASAIO J. 2015;61:540–7.View ArticleGoogle Scholar
- Rausch MK, Tibayan FA, Ingels NB, Miller DC, Kuhl E. Mechanics of the mitral annulus in chronic ischemic cardiomyopathy. Ann Biomed Eng. 2013;41:2171–80.View ArticleGoogle Scholar
- Nakai H, Kaku K, Takeuchi M, Otani K, Yoshitani H, Haruki N, et al. Different influences of left ventricular remodeling on anterior and posterior mitral leaflet tethering. Circ J. 2012;76:2481–7.View ArticleGoogle Scholar
- Mann DL, Bristow MR. Mechanisms and models in heart failure: the biomechanical model and beyond. Circulation. 2005;111:2837–49.View ArticleGoogle Scholar
- Vedula V, Seo JH, Lardo AC, Mittal R. Effect of trabeculae and papillary muscles on the hemodynamics of the left ventricle. Theor Comput Fluid Dyn. 2015;30(1):3–21.Google Scholar