Académique Documents
Professionnel Documents
Culture Documents
com
Heart Online First, published on October 28, 2015 as 10.1136/heartjnl-2015-308044
Review
Review
Review
Table 1 Continued
Area Clinical applications Data and evidence Potential clinical impact Limitations and challenges References
vascular disease phenotype, vascular diseases, including progression of WSS map. Acquisition of
is contributing to the atherosclerosis, aneurysm atherosclerosis. WSS map patient specific boundary
understanding of cellular and post-stent NH combined with multi-scale conditions remains clinically
biology modelling may inform clinical challenging.
practice, such as the site of
rupture in aneurysm, and
severity of in-stent restenosis.
2 34 35 36 37 38 39 40 41
Heart failure Models based upon CT and CFD/FSI models replicate Additional haemodynamic Resolution of imaging and
MR help compute realistic pathophysiology in data potentially enables early reconstruction (representing
haemodynamics and the models of health and diagnosis and stratifies trabeculae and papillary
spatio-temporal distributions disease (eg, HFREF, HFPEF, disease phenotypes and muscles). Tuning with realistic
of pressure and myocardial HCM, DCM, and RWMA severities. Characterising boundary conditions.
stress/strain post-MI) complex vortex flows Requirement for FSI in many
identifies areas of flow models
stagnation and thrombus risk
CRT Coupled electro-mechanical Published reports of Improved selection of CRT Uncertainties and
models of the ventricle accurate patient-specific responders. Simulation and assumptions regarding
incorporating CFD haemodynamic simulations selection of optimal tuning of boundary conditions and the
(multi-physics models) used with sufficient detail to device settings and lead range of clinical
to investigate heart function optimise CRT before placement on an individual measurements required for
surgical intervention case basis parameterisation. Mesh
generation, prolonged
computation times
VADs Generic optimisation of Published models describing Pump tuning to ensure Post-implantation imaging
pump design. haemodynamic influences of periodic opening and closing artefact limits modelling.
Patient-specific models can catheter placement and of AV, preventing leaflet Optimising performance
aid implantation strategy minimisation of adverse fusion. Personalised catheter requires the balance of
and tuning of output haemodynamic effects placement planning multiple competing factors.
according to patient (prediction and avoidance As for all cardiac
physiology stasis and thrombus electromechanical models,
formation) selection of appropriate
patient specific parameters is
difficult due to sparsity of
data
42 43 44
Congenital heart CFD simulates Range of models described, Modelling enables greater Acquisition and application of
disease haemodynamics which are including reduced order, 3D understanding of systemic model parameters and
complex and hard to predict CFD, FSI and multiscale, and regional haemodynamics boundary conditions from
in the context of a diverse particularly in the context of and the prediction of patient and literature data.
and heterogeneous range of univentricular circulation, response to putative surgical The ultimate personalisation
disease phenotypes aortic and pulmonary or device-based treatments challenge
malformations which often involve
significant modifications to
the circulatory tree
AV, aortic valve; CFD, computational fluid dynamics; CRT, cardiac resynchronisation therapy; CT (A), CT (angiography); DCM, dilated cardiomyopathy; FSI, fluid solid interaction; HCM,
hypertrophic cardiomyopathy; HFPEF, heart failure with preserved EF; HFREF, heart failure with reduced EF; MI, myocardial infarction; NH, neointimal hyperplasia; PAP, pulmonary artery
pressure; PC, phase-contrast; PCI, percutaneous coronary intervention; RA, regulatory authority; RWMA, regional wall motion abnormality; (v)FFR; (virtual) fractional flow reserve; VAD,
ventricular assist device; WSS, wall shear stress.
MODEL CONSTRUCTION both spatial and temporal renement. The fabrication of the
CFD model construction and solution can be described in seven mesh, and the level of mesh renement, are inuenced by
stages (gure 1): case- and context-specic factors. The mesh and timestep (ie,
1. Clinical imaging spatio-temporal discretisation) must be rened enough to
A range of medical imaging modalities can be used, including capture the important haemodynamic behaviour of the mod-
ultrasound, CT, MRI and X-ray angiography. Imaging must elled compartment (the nal solution should be independent of
provide sufcient anatomical and physiological detail, in an mesh parameters), but without excessive renement because this
appropriate format and quality, to enable segmentation and data impacts negatively on computational resource and solution time
extraction.49 (see online supplementary table S1).
2. Segmentation and reconstruction 4. Boundary conditions
Segmentation methods convert medical images to in silico Because it is impossible to discretise the entire cardiovascular
geometries which dene the physical bounds of the model system, the region to be analysed will have at least one inlet and
region of interest. If images are acquired over a cardiac cycle, one outlet. To enable CFD analysis, the physiological conditions
anatomical motion can be tracked over segmented regions.50 51 at the wall and inlet/outlet boundaries must be specied.
3. Discretisation Boundary conditions are a set of applied physiological para-
Spatial discretisation, or meshing, divides the geometry into meters (which may vary over time) that dene the physical con-
a number of discrete volumetric elements or cells. Temporal dis- ditions at the inlets, outlets and walls. They may be based on
cretisation divides the solution into discrete time steps. The patient-specic data, population data, physical models or
accuracy and numerical stability of the analysis are inuenced by assumptions.39
Morris PD, et al. Heart 2015;0:111. doi:10.1136/heartjnl-2015-308044 3
Downloaded from http://heart.bmj.com/ on May 28, 2017 - Published by group.bmj.com
Review
Review
Table 3 A summary of the various orders of CFD modelling applied to the cardiovascular system
Typical solution
Model Figure CFD solution Description/examples time
0D * No spatial dimension. Physiological Lump together distributed physiological systems Immediate solution
variables such as pressure (P), flow (Q) into a single description. They describe the global
and resistance (R) are assumed behaviour of the modelled segment. The 0D
spatially uniform within the model, Windkessel model (pictured) is often used to
varying only as a function of time (t), represent the compliant and resistive nature of
eg, the arterial circulation. 0D models are frequently
Pt Qt Rt used to model components of the cardiovascular
Solved with ordinary differential (0D system or to improve boundary conditions for 3D
NS) equations models of arterial, ventricular or venous
pathophysiology.5 47
1D Physiological variables are solved as a Used to represent wave propagation S (static)
function of a single spatial variable, characteristics and wave reflection. 1D models Min (transient)
typically length (x), eg, may also be used to provide boundary conditions
Px;t Qx;t Rx;t for higher order models in order to increase
Solved with partial differential (1D NS) refinement of the solution, especially where the
equations effects of wave reflection are significant.45 46
2D Physiological variables are solved as a Able to resolve the solution in 2D. Used less
function of two spatial variables, often now than previously due to ready
typically length and distance from availability of improved computer processing and
centreline (r) eg, 3D solvers. Examples include the simulation of
Px;r;t Qx;r;t Rx;r;t para-prosthetic valve haemolysis and
Solved with axisymmetric NS improvement of the assessment of the proximal
equations flow convergence zone in the clinical evaluation
of regurgitant valve disease.48
3D Physiological variables are solved as a Full 3D CFD can resolve the physiological solution Order of minutes for
function of all three spatial variables, in all dimensions including time. Examples are steady- state
including the angle around the more widely reviewed in the main body of the Order of hours or
centreline axis () eg, text. days for transient
Px;r;u;t Qx;r;u;t Rx;r;u;t
Solved with full 3-D NS equations
*Hydro-electrical analogue diagrams are often used to describe physiological components such as resistance, pressure (voltage), compliance (capacitance), and flow (current).
Solution times vary according to complexity of the model and the mathematical solution. The times presented are approximate and are based on a model of coronary physiology.5
CFD, computational fluid dynamics; NS, Navier-Stokes; 0D, zero dimensional; 1D, one dimensional; 2D, two dimensional; 3D, three dimensional.
approximation is acceptable for some applications.53 54 Vessel Typically, cardiovascular simulations assume blood behaves as an
compliance allows blood to be stored during systole and incompressible uid. Although blood exhibits non-Newtonian
released during diastole. At the system level this results in a behaviour (see glossary in table 2), which must be simulated for
nite speed of pressure wave transmission and tends to reduce ow in small capillaries, in larger vessels these effects are often
the peak pressures associated with the inertial acceleration of neglected and a Newtonian uid model is assumed.
the blood. Compliance tends to reduce shear stresses because
the vessels are slightly larger when peak ow occurs. It is pos- BENEFITS OF CARDIOVASCULAR CFD MODELLING
sible to model deformation of the wall, due to cardiac and CFD modelling enables investigation of pressure and ow elds
respiratory variation, in response to change in pressure using at a temporal and spatial resolution unachievable by any clinical
uid-solid interaction (FSI) models. These are far more complex methodology. Post-processing provides additional data, generat-
to solve, boundary conditions are a challenge, and many wall ing new insights into physiology and disease processes. For
parameters remain unknown, which increases the number of example, it is difcult, and invasive, to measure arterial wall
assumptions. Furthermore, it is yet to be established for which shear stress (WSS), a key factor in the development of athero-
applications a full FSI approach improves accuracy. An example sclerosis and in-stent restenosis, whereas CFD models can
of the increased computational cost of FSI is reported by Brown compute WSS and map its spatial distribution.20 21 Such work
et al54 where 3D transient (time-varying) analysis of the aorta has established the link between haemodynamic disturbance and
required 145 h (FSI) compared with 6.6 h (CFD). An alternative atherogenesis and has explained the preferential deposition of
is to impose wall movement derived from imaging data (eg, atherosclerotic plaque at arterial bends and bifurcation
gated MRI). There are exciting developments in the use of data regions.22 CFD modelling has been central to our current
assimilation techniques in which sparse clinical data, for understanding of the effects of WSS on endothelial homeostasis:
example, from 4D imaging, are integrated with the analysis laminar, non-disturbed blood ow is associated with increased
process so that material properties of tissues in individual WSS which inhibits unnecessary endothelial cell activation;
patients are recovered as the simulation progresses.55 In biomed- whereas turbulent or disturbed blood ow reduces WSS which
ical workows it is assumed the boundaries of the uid geom- stimulates adverse vessel remodelling. A complex series of
etry are smooth, yet medical images may not generate smooth WSS-related signalling pathways and interactions underlie this
surfaces due to poor resolution or imaging artefacts. Instead, phenomenon. Before these pathways can be exploited to gener-
structures may be smoothed in silico after segmentation. ate anti-atherosclerotic therapies their complexity needs to be
Morris PD, et al. Heart 2015;0:111. doi:10.1136/heartjnl-2015-308044 5
Downloaded from http://heart.bmj.com/ on May 28, 2017 - Published by group.bmj.com
Review
Review
Review
Figure 3 A computational uid dynamics (CFD) model demonstrating the correlation between wall shear stress (WSS) and restenosis in coronary
artery disease. (A) Structural modelling of stent insertion in porcine coronary arteries reconstructed from micro-CT, and stentartery coupling
obtained after arterial recoil. (B) Comparison between the in vivo histological images (left) and corresponding sections from the structural simulation
(right) demonstrating excellent agreement. (C) Results of the CFD simulations in terms of the spatial distribution of WSS magnitude over the arterial
wall. (D) The correlation between areas characterised by low WSS (orange lines) and in-stent restenosis after 14 days. The CFD simulation of WSS
has identied areas of reduced shear and restenosis with excellent agreement. Figure reproduced from Morlacchi et al26 with kind permission from
Springer Science and Business Media.
dissections remains controversial. Thoracic endovascular aortic impact of various conditions, pathologies and treatments across
repair (TEVAR) carries a risk of spinal ischaemia, and multiple multiple organ systems. While this is ambitious, large, inter-
communicating channels between true and false lumens confer a national, collaborative research projects under the umbrella of
risk of proximal rupture if only the primary entry is closed. the Virtual Physiological Human (VPH) reect the seriousness,
Therefore, graft length is balanced against the risk of paraplegia. legitimacy and rationale underlying this long-term vision.68 Full
In this context, CFD modelling provides individualised risk system models offer the possibility of understanding, holistically,
stratication and optimised treatment delivery (gure 5).65 18 the impact of cardiovascular disease upon individual patients.
Similar simulations are useful in the context of abdominal aortic
aneurysm.66 67 IN SILICO TRIALS
In silico techniques can simulate measures of safety, accuracy,
FULL SYSTEM MODELS: THE VIRTUAL PHYSIOLOGICAL and efcacy of interventions, both pharmacological and mech-
HUMAN anical, in large cohorts of virtual patient models which represent
There is increasing interest in integrating multiple physiological naturally occurring physiological and pathological variability.
models into comprehensive system models to simulate the This minimises the time, cost, and risk associated with clinical
Figure 4 Computational uid dynamics (CFD) model of an intracranial berry aneurysm from the @neurist project. Panel (A) demonstrates the
reconstructed surface mesh. Panels (B) and (C) demonstrate the CFD simulated pressure (B) and wall shear stress (C) acting upon the aneurysm wall,
which may be useful in predicting risk of rupture on a patient-specic basis.
8 Morris PD, et al. Heart 2015;0:111. doi:10.1136/heartjnl-2015-308044
Downloaded from http://heart.bmj.com/ on May 28, 2017 - Published by group.bmj.com
Review
Figure 5 Segmentation, reconstruction and 3D simulation of a chronic type B aortic dissection with true and false lumen in systole (top row) and
diastole (bottom row). The primary entry point (top arrow) is close to the left subclavian artery. Two more communications (re-entries) are seen
distally. Computational uid dynamics simulation allows the ow through each re-entry point to be studied separately in order to predict response
to intervention. During systole simulation demonstrates high blood ow velocity through the primary entry point. However, simulation predicts
signicant ow through the rst re-entry point in systole, and even higher during diastole, thus demonstrating that closure of the primary entry
point alone will not be sufcient to induce false lumen thrombosis and avoid further expansion. Reproduced with permission from Chen et al,
2013.18
trials. The Avicenna project leverages signicant commercial Model accuracy is determined by model design and quality of
interest and engagement to develop a roadmap describing the input data. For CFD applications it is unclear how detailed the
route by which multi-scale in silico techniques can achieve clinical data needs to be in terms of geometry (segmented from
this.69 medical images) and parameterisation (variability described by
the model and the tuning of patient-specic boundary condi-
CHALLENGES AND LIMITATIONS tions). Continuing improvements in imaging, image-registration
CFD models in medicine have traditionally been used by two and segmentation algorithms will augment accuracy.50 Model
user groups: industrial medical device developers, in rapid, parameterisation is more challenging, because it requires
low-cost, device prototyping; and academics, to investigate car- detailed knowledge of physiological metrics in the proximal and
diovascular physiology and compute parameters that cannot distal circulations which may be inaccessible and variable in
otherwise be obtained. Both groups construct models which are health and in disease, for example, microcirculatory resistance is
typically complex, involve multiple nely-tuned geometric, a major determinant of coronary blood ow. Further under-
haemodynamic and material parameters and require long com- standing of the relative importance of physiological parameters
putation times. In contrast, clinicians are a third, emerging user is required to determine those which are most inuential, and
group, who require rapid results with adequate accuracy. those which can be assumed or averaged.70 This allows
Morris PD, et al. Heart 2015;0:111. doi:10.1136/heartjnl-2015-308044 9
Downloaded from http://heart.bmj.com/ on May 28, 2017 - Published by group.bmj.com
Review
unnecessary model complexity to be simplied, balancing com- 4 Radaelli AG, Augsburger L, Cebral JR, et al. Reproducibility of haemodynamical
puting speed against accuracy.54 A second challenge is the devel- simulations in a subject-specic stented aneurysm modela report on the Virtual
Intracranial Stenting Challenge 2007. J Biomech 2008;41:206981.
opment of relevant industry standards. In the European Union 5 Morris PD, Ryan D, Morton AC, et al. Virtual fractional ow reserve from coronary
and USA, diagnostic software is regulated via CE marking and angiography: modeling the signicance of coronary lesions: results from the
FDA directives, respectively; but there are no industry standards VIRTU-1 (VIRTUal Fractional Flow Reserve From Coronary Angiography) Study. JACC
governing accuracy, reliability or validation. The FDA is addres- Cardiovasc Interv 2013;6:14957.
6 Nrgaard BL, Leipsic J, Gaur S, et al. Diagnostic performance of noninvasive
sing this through benchmarking initiatives, in the same way that
fractional ow reserve derived from coronary computed tomography angiography in
aviation authorities adopted computer-aided design over trad- suspected coronary artery disease: the NXT trial (Analysis of Coronary Blood Flow
itional physical testing.71 Third, large volumes of clinical data, Using CT Angiography: Next Steps). J Am Coll Cardiol 2014;63:114555.
of value for model development and validation, are stored in 7 Tu S, Barbato E, Kszegi Z, et al. Fractional ow reserve calculation from
hospital systems. Access is variable and restrictive; although the 3-dimensional quantitative coronary angiography and TIMI frame count: a fast
computer model to quantify the functional signicance of moderately obstructed
VPH-Share project demonstrates how anonymised patient- coronary arteries. JACC Cardiovasc Interv 2014;7:76877.
specic data and in silico models can be shared securely.72 8 Morris PD, van de Vosse FN, Lawford PV, et al. Virtual (computed) fractional
Access to such data supports model validation against long-term ow reserve: current challenges and limitations. JACC Cardiovasc Interv
outcomes which may expedite clinical translation. Fourth, CFD 2015;8:100917.
9 Sotiropoulos F, Borazjani I. A review of state-of-the-art numerical methods for
modelling can be perceived as a disruptive technology, and a
simulating ow through mechanical heart valves. Med Biol Eng Comput
threat, especially by manufacturers of traditional hardware. 2009;47:24556.
Finally, the next generation of doctors will require training in in 10 Simon HA, Ge L, Borazjani I, et al. Simulation of the three-dimensional hinge ow
silico systems, and understanding their principles, methodolo- elds of a bileaet mechanical heart valve under aortic conditions. Ann Biomed Eng
gies and limitations. Several initiatives are attempting to address 2010;38:84153.
11 Yoganathan AP, Chandran KB, Sotiropoulos F. Flow in prosthetic heart valves:
this, but in the longer term, in silico medicine may gain promin- state-of-the-art and future directions. Ann Biomed Eng 2005;33:168994.
ence in medical curricula.68 73 74 12 Astorino M, Hamers J, Shadden SC, et al. A robust and efcient valve model based
on resistive immersed surfaces. Int J Numer Method Biomed Eng 2012;28:93759.
13 De Hart J, Peters GW, Schreurs PJ, et al. A three-dimensional computational
FUTURE DIRECTIONS analysis of uid-structure interaction in the aortic valve. J Biomech
2003;36:10312.
In the context of device development, the major computational 14 Erhart P, Hyhlik-Drr A, Geisbsch P, et al. Finite element analysis in asymptomatic,
modelling and technological challenges have already been symptomatic, and ruptured abdominal aortic aneurysms: in search of new rupture
addressed and the benets of simulation are recognised by the risk predictors. Eur J Vasc Endovasc Surg 2015;49:23945.
regulatory authorities. Sustained investment will enable engi- 15 Georgakarakos E, Ioannou CV, Papaharilaou Y, et al. Computational evaluation of
neers to continue model renement and the development of aortic aneurysm rupture risk: what have we learned so far? J Endovasc Ther
2011;18:21425.
novel applications, specically targeting increases in precision, 16 Molony DS, Kavanagh EG, Madhavan P, et al. A computational study of the
personalisation and speed. Beyond technological development, magnitude and direction of migration forces in patient-specic abdominal aortic
and before these tools become established in routine clinical aneurysm stent-grafts. Eur J Vasc Endovasc Surg 2010;40:3329.
practice, the most immediate need is to demonstrate equivalence 17 Karmonik C, Mller-Eschner M, Partovi S, et al. Computational uid dynamics
investigation of chronic aortic dissection hemodynamics versus normal aorta. Vasc
of in silico results relative to invasive measurements through Endovascular Surg 2013;47:62531.
observational trials. Beyond this, efcacy must be demonstrated 18 Chen D, Mller-Eschner M, von Tengg-Kobligk H, et al. A patient-specic study of
in large multicentre clinical trials. It is clear that these techni- type-B aortic dissection: evaluation of true-false lumen blood exchange. Biomed
ques have the potential to change clinical practice. The bene- Eng Online 2013;12:65.
ciaries will be patients, clinicians and healthcare providers. 19 Cheng Z, Juli C, Wood NB, et al. Predicting ow in aortic dissection: comparison of
computational model with PC-MRI velocity measurements. Med Eng Phys
Acknowledgements Dr Morris is funded by a Clinical Research Training 2014;36:117684.
Fellowship from the British Heart Foundation. Figure 4 was kindly provided by 20 Chien S. Mechanotransduction and endothelial cell homeostasis: the wisdom of the
Dr Alberto Marzo, University of Shefeld. cell. Am J Physiol Heart Circ Physiol 2007;292:H120924.
21 LaDisa JF, Jr., Olson LE, Molthen RC, et al. Alterations in wall shear stress predict
Contributors All authors have made a substantial contribution to the conception sites of neointimal hyperplasia after stent implantation in rabbit iliac arteries. Am J
or design of the work, the drafting and revision of the work and have made an Physiol Heart Circ Physiol 2005;288:H246575.
intellectual contribution. All authors have given nal approval of the version 22 Jin S, Yang Y, Oshinski J, et al. Flow patterns and wall shear stress distributions at
submitted and have agreed to be accountable for all aspects of the work. atherosclerotic-prone sites in a human left coronary arteryan exploration using
Funding British Heart Foundation (R/134747-11-1). combined methods of CT and computational uid dynamics. Conf Proc IEEE Eng
Med Biol Soc 2004;5:378991.
Competing interests None declared. 23 Lee J, Smith NP. The multi-scale modelling of coronary blood ow. Ann Biomed Eng
Provenance and peer review Not commissioned; externally peer reviewed. 2012;40:2399413.
24 Jimnez JM, Davies PF. Hemodynamically driven stent strut design. Ann Biomed Eng
Open Access This is an Open Access article distributed in accordance with the 2009;37:148394.
terms of the Creative Commons Attribution (CC BY 4.0) license, which permits 25 Seshadhri S, Janiga G, Beuing O, et al. Impact of stents and ow diverters on
others to distribute, remix, adapt and build upon this work, for commercial use, hemodynamics in idealized aneurysm models. J Biomech Eng 2011;133:071005.
provided the original work is properly cited. See: http://creativecommons.org/ 26 Morlacchi S, Migliavacca F. Modeling stented coronary arteries: where we are,
licenses/by/4.0/ where to go. Ann Biomed Eng 2013;41:142844.
27 Wentzel JJ, Krams R, Schuurbiers JC, et al. Relationship between neointimal
thickness and shear stress after Wallstent implantation in human coronary arteries.
REFERENCES Circulation 2001;103:17405.
1 Taylor CA, Figueroa CA. Patient-specic modeling of cardiovascular mechanics. 28 Peach TW, Ngoepe M, Spranger K, et al. Personalizing ow-diverter intervention for
Annu Rev Biomed Eng 2009;11:10934. cerebral aneurysms: from computational hemodynamics to biochemical modeling.
2 Smith N, de Vecchi A, McCormick M, et al. euHeart: personalized and integrated Int J Numer Method Biomed Eng 2014;30:1387407.
cardiac care using patient-specic cardiovascular modelling. Interface Focus 29 Schneiders JJ, Marquering HA, van Ooij P, et al. Additional value of
2011;1:34964. intra-aneurysmal hemodynamics in discriminating ruptured versus unruptured
3 Gasser TC, Nchimi A, Swedenborg J, et al. A novel strategy to translate the intracranial aneurysms. AJNR Am J Neuroradiol 2015;10:19206.
biomechanical rupture risk of abdominal aortic aneurysms to their equivalent 30 Qureshi MU, Vaughan GA, Sainsbury C, et al. Numerical simulation of blood ow
diameter risk: method and retrospective validation. Eur J Vasc Endovasc Surg and pressure drop in the pulmonary arterial and venous circulation. Biomech Model
2014;47:28895. Mechanobiol 2014;13:113754.
Review
31 Lungu A, Wild JM, Capener D, et al. MRI model-based non-invasive differential 53 Zeng D, Boutsianis E, Ammann M, et al. A study on the compliance of a right
diagnosis in pulmonary hypertension. J Biomech 2014;47:29417. coronary artery and its impact on wall shear stress. J Biomech Eng
32 Tang BT, Pickard SS, Chan FP, et al. Wall shear stress is decreased in the pulmonary 2008;130:041014.
arteries of patients with pulmonary arterial hypertension: an image-based, 54 Brown AG, Shi Y, Marzo A, et al. Accuracy vs. computational time: translating
computational uid dynamics study. Pulm Circ 2012;2:4706. aortic simulations to the clinic. J Biomech 2012;45:51623.
33 Kheyfets VO, Rios L, Smith T, et al. Patient-specic computational modeling of 55 Bertoglio C, Barber D, Gaddum N, et al. Identication of artery wall stiffness: in
blood ow in the pulmonary arterial circulation. Comput Methods Programs Biomed vitro validation and in vivo results of a data assimilation procedure applied to a 3D
2015;120:88101. uid-structure interaction model. J Biomech 2014;47:102734.
34 Niederer SA, Plank G, Chinchapatnam P, et al. Length-dependent tension in the 56 Reporting of Computational Modeling Studies in Medical Device Submissions.
failing heart and the efcacy of cardiac resynchronization therapy. Cardiovasc Res Rockeville, MD: Draft Guidance for Industry and Food and Drug Administration
2011;89:33643. Staff, 2014.
35 Chan BT, Lim E, Chee KH, et al. Review on CFD simulation in heart with dilated 57 Bluestein D, Girdhar G, Einav S, et al. Device thrombogenicity emulation: a novel
cardiomyopathy and myocardial infarction. Comput Biol Med 2013;43:37785. methodology for optimizing the thromboresistance of cardiovascular devices.
36 Aguado-Sierra J, Krishnamurthy A, Villongco C, et al. Patient-specic modeling of J Biomech 2013;46:33844.
dyssynchronous heart failure: a case study. Prog Biophys Mol Biol 2011;107:14755. 58 Pijls NH, Fearon WF, Tonino PA, et al. Fractional ow reserve versus angiography
37 Sermesant M, Chabiniok R, Chinchapatnam P, et al. Patient-specic for guiding percutaneous coronary intervention in patients with multivessel coronary
electromechanical models of the heart for the prediction of pacing acute effects in artery disease: 2-year follow-up of the FAME (Fractional Flow Reserve Versus
CRT: a preliminary clinical validation. Med Image Anal 2012;16:20115. Angiography for Multivessel Evaluation) study. J Am Coll Cardiol 2010;56:17784.
38 Brown AG. Patient-specic local and systemic haemodynamics in the presence of a 59 Papafaklis MI, Muramatsu T, Ishibashi Y, et al. Fast virtual functional assessment of
left ventricular assist device. Department of cardiovascular science. Shefeld: intermediate coronary lesions using routine angiographic data and blood ow
University of Shefeld, 2012. simulation in humans: comparison with pressure wirefractional ow reserve.
39 Brown AG, Shi Y, Arndt A, et al. Importance of realistic LVAD proles for assisted EuroIntervention 2014;10:57483.
aortic simulations: evaluation of optimal outow anastomosis locations. Comput 60 U.S. Food and Drug Administration. De novo classication request for FFRCT V. 1.4
Methods Biomech Biomed Engin 2012;15:66980. (De Novo Summary (DEN130045)). 2014.
40 Chiu WC, Girdhar G, Xenos M, et al. Thromboresistance comparison of the 61 Sonntag SJ, Li W, Becker M, et al. Combined computational and experimental
HeartMate II ventricular assist device with the device thrombogenicity emulation- approach to improve the assessment of mitral regurgitation by echocardiography.
optimized HeartAssist 5 VAD. J Biomech Eng 2014;136:021014. Ann Biomed Eng 2014;42:97185.
41 Farag MB, Karmonik C, Rengier F, et al. Review of recent results using 62 Swift AJ, Wild JM, Nagle SK, et al. Quantitative magnetic resonance imaging of
computational uid dynamics simulations in patients receiving mechanical assist pulmonary hypertension: a practical approach to the current state of the art.
devices for end-stage heart failure. Methodist Debakey Cardiovasc J 2014;10:1859. J Thorac Imaging 2014;29:6879.
42 Ladisa JF Jr, Taylor CA, Feinstein JA. Aortic coarctation: recent developments in 63 Morlacchi S, Colleoni SG, Crdenes R, et al. Patient-specic simulations of stenting
experimental and computational methods to assess treatments for this simple procedures in coronary bifurcations: two clinical cases. Med Eng Phys
condition. Prog Pediatr Cardiol 2010;30:459. 2013;35:127281.
43 Pennati G, Corsini C, Hsia TY, et al. Computational uid dynamics models and 64 Villa-Uriol MC, Berti G, Hose DR, et al. @neurIST complex information processing
congenital heart diseases. Front Pediatr 2013;1:4. toolchain for the integrated management of cerebral aneurysms. Interface Focus
44 Pittaccio S, Migliavacca F, Dubini G, et al. On the use of computational models for 2011;1:30819.
the quantitative assessment of surgery in congenital heart disease. Anadolu Kardiyol 65 Chen D, Mller-Eschner M, Kotelis D, et al. A longitudinal study of type-B aortic
Derg 2005;5:2029. dissection and endovascular repair scenarios: computational analyses. Med Eng
45 van de Vosse FN, Stergiopulos N. Pulse wave propagation in the arterial tree. Annu Phys 2013;35:132130.
Rev Fluid Mech 2011;43:46799. 66 Venkatasubramaniam AK, Fagan MJ, Mehta T, et al. A comparative study of aortic
46 Shi Y, Lawford P, Hose R. Review of zero-D and 1-D models of blood ow in the wall stress using nite element analysis for ruptured and non-ruptured abdominal
cardiovascular system. Biomed Eng Online 2011;10:33. aortic aneurysms. Eur J Vasc Endovasc Surg 2004;28:16876.
47 Kim HJ, Vignon-Clementel IE, Figueroa CA, et al. On coupling a lumped parameter 67 Martu G, Christian Gasser T. Review: the role of biomechanical modeling in the
heart model and a three-dimensional nite element aorta model. Ann Biomed Eng rupture risk assessment for abdominal aortic aneurysms. J Biomech Eng
2009;37:215369. 2013;135:021010.
48 Rodriguez L, Thomas JD, Monterroso V, et al. Validation of the proximal ow 68 Hunter P, Chapman T, Coveney PV, et al. A vision and strategy for the virtual
convergence method. Calculation of orice area in patients with mitral stenosis. physiological human: 2012 update. Interface Focus 2013;3:20130004.
Circulation 1993;88:115765. 69 http://www.avicenna-isct.org (accessed Jul 2015).
49 Weese J, Groth A, Nickisch H, et al. Generating anatomical models of the heart and 70 Donders WP, Huberts W, van de Vosse FN, et al. Personalization of models with
the aorta from medical images for personalized physiological simulations. Med Biol many model parameters: an efcient sensitivity analysis approach. Int J Numer
Eng Comput 2013;51:120919. Method Biomed Eng 2015;31. Published Online First: 27th May 2015.
50 Barber DC, Hose DR. Automatic segmentation of medical images using image doi:10.1002/cnm.2727. Epub 2015 Jun 15
registration: diagnostic and simulation applications. J Med Eng Technol 71 Stewart S, Hariharan P, Paterson E, et al. Results of FDAs rst interlaboratory
2005;29:5363. computational study of a nozzle with a sudden contraction and conical diffuser.
51 Zhao F, Xie X. An overview of interactive medical image segmentation. Ann BMVA Cardiovasc Eng Technol 2013;4:37491.
2013;7:122. 72 vph-share.eu. http://vph-share.eu (accessed Jul 2015).
52 Kung E, Pennati G, Migliavacca F, et al. A simulation protocol for exercise 73 Cates CU, Gallagher AG. The future of simulation technologies for complex
physiology in Fontan patients using a closed loop lumped-parameter model. cardiovascular procedures. Eur Heart J 2012;33:212734.
J Biomech Eng 2014;136:107001. 74 http://vph-portal.eu (accessed Jul 2015).
These include:
Notes