Influence of surface model extraction parameter on computational fluid dynamics modeling of cerebral aneurysms
Introduction
Because of the postulated relationship between hemodynamic factors and the natural history of cerebral aneurysms, numerous investigators have studied intra-aneurysmal blood flow and wall shear stress (WSS) patterns. WSS was reported as a possible indicator of aneurysm initiation, growth, and rupture, (Malek and Izumo, 1995, Metaxa et al., 2010). However, there are two opposing hypothesis regarding growth and rupture, i.e., that high (Cebral et al., 2011, Cebral et al., 2010, Hassan et al., 2005, Hoi et al., 2004) or low WSS (Boussel et al., 2008, Doenitz et al., 2010, Jou et al., 2008, Metaxa et al., 2010) is responsible for the growth and/or rupture of aneurysm.
While early studies investigated hemodynamics in idealized arterial geometries, the current trend is to use realistic replicas of in vivo arteries. It is mandatory to adopt anatomically realistic geometrical models on an individual basis, since individual variability in arterial configurations is the rule rather than the exception, and blood flow characteristics strongly depend on vessel configuration (Cebral and Löhner, 2001). With this approach, 3-dimensional rotational angiography (3DRA) is used to obtain patient-specific vascular models.
Commercially available software packages are now available for simplify patient-specific modeling (Metcalfe, 2003), but determining the threshold image intensity in reconstructing a vascular model from medical imaging data, which involves lumen boundary identification, can affect the configuration of a vascular model and the results of numerical simulation. Although some researchers have established automated methods for vascular model reconstruction (Antiga et al., 2008, Cebral et al., 2005, Chang et al., 2009, Yim et al., 2002), the threshold image intensity is still generally determined by trial-and-error and visual inspection (Chang et al., 2009, Hassan et al., 2004, Rayz et al., 2008, Venugopal et al., 2007). With the trial-and-error based method, accuracy and reproducibility depend on the skills of users, and numerical simulation based on this method may not accurately reproduce intra-aneurysmal hemodynamics of an individual patient's aneurysm. A minimally subjective threshold determination method is thus required.
We created a new method of threshold determination and reconstructed patient-specific vascular models employing various threshold image intensity determined by this method. Then, we assessed the effects of threshold image intensity differences on vascular model configuration and the WSS distribution of the aneurysm.
Section snippets
Subjects
This study included consecutive 11 patients (8 females and 3 males) 43–77 years of age (median, 68 years) with internal carotid artery aneurysms. These 11 patients underwent open surgery between April 2009 and March 2010. Pre-operative 3DRA was performed in all patients. We included aneurysms with a maximum diameter of more than 5 mm. The clinical characteristics are summarized in Table 1. The study protocol was approved by the local ethics committee. All patients gave written informed consent
Results
Forty-four vascular models of 11 aneurysms were created according to the four Cthre values, and numerical simulation could be conducted for all models.
In the threshold image intensity determination, the highest and baseline values were 4820 (4090–6440) and 340 (−130–700), respectively. Threshold image intensity calculated according to Cthre values of 0.3, 0.4, 0.5 and 0.6 were 1850 (1180–2250), 2340 (1610–2810), 2750 (2050–3420) and 3170 (2480–4020), respectively. Vascular models of all 11
Discussion
A successful CFD study relies on an accurate description of the vascular configuration that is obtained from volumetric images. Our findings show that threshold image intensity differences in reconstructing vascular models affect the configuration of these vascular models and the WSS distribution of the aneurysm.
Conclusions
Threshold image intensity differences can profoundly affect CFD. Our results suggest the importance of the procedure used for threshold determination, and the uniform setting of the Cthre value is important for an objective CFD. More extensive and detailed investigations using objective threshold determination method are necessary to gain a further understanding of problems in image segmentation, and to simplify the application of patient-specific CFD modeling of cerebral aneurysms.
Conflict of interests
Each of the authors does not have any conflict of interest.
Acknowledgments
There are no other contributors, and the authors acknowledge no funding sources or granting agencies that supported this work.
References (27)
- et al.
Control of endothelial cell gene expression by flow
Journal of Biomechanics
(1995) - et al.
Computational blood flow modelling: errors associated with reconstructing finite element models from magnetic resonance images
Journal of Biomechanics
(1997) - et al.
A calculation procedure for heat, mass and momentum transfer in three-dimensional parabolic flows
International Journal of Heat and Mass Transfer
(1972) - et al.
The importance of parent artery geometry in intra-aneurysmal hemodynamics
Medical Engineering and Physics
(2008) - et al.
An image-based modeling framework for patient-specific computational hemodynamics
Medical and Biological Engineering and Computing
(2008) - et al.
Aneurysm growth occurs at region of low wall shear stress
Stroke
(2008) - et al.
Computational fluid dynamics modeling of intracranial aneurysms: effects of parent artery segmentation on intra-aneurysmal hemodynamics
American Journal of Neuroradiology
(2006) - et al.
Hemodynamic patterns of anterior communicating artery aneurysms: a possible association with rupture
American Journal of Neuroradiology
(2009) - et al.
Efficient pipeline for image-based patient-specific analysis of cerebral aneurysm hemodynamics: technique and sensitivity
IEEE Transactions on Medical Imaging
(2005) - et al.
From medical images to anatomically accurate finite element grids
International Journal for Numerical Methods in Engineering
(2001)