International Journal of Computational Methods

Paravalvular aortic regurgitation (AR) significantly influences the outcome of transcatheter aortic valve implantation (TAVI). This study deals with the effect of stent design on AR after TAVI was evaluated by numerical simulations. Three different stents were implanted into the representative calcified human aortic root models. The excessively dense or a sparse outflow end would result in larger stent deformation indexes, higher stress of calcified plaques and aortic root, and larger gap between stent and aortic wall, and lead to serious paravalvular regurgitation. This study provides the guidance for design of transcatheter aortic valve.

Instantaneous wave-free ratio (iFR), an invasive index of coronary artery tree, can evaluate the functional performance of vascular stenosis without pharmacological vasodilators. The noninvasive assessment of diameter stenosis (DS) obtained from coronary computed tomography angiography (CTA) has high false positive rate in contrast to iFR. The aim of this study was to develop a numerical simulation method that predicts the iFR and noninvasively assess the myocardial ischemia. Based on the CTA images, a patient-specific three-dimensional model of the aorta and coronary arteries were reconstructed. A stenosis was created in the left anterior descending artery (LAD) by reducing the DS of geometric model (40%, 50%, 60%, 75% and 90%). The patient-specific LPM boundary condition were set up to compute iFRct value during the wave-free period at the resting condition. The computed pressure and flow of coronary artery were realistic as compared to literature data. In contrast to invasive iFR, the iFRct can make a cost-benefit balance in terms of clinical cost and patient’s health.

The stents’ adaptability and safety in realistic and idealized stenotic coronary model were compared to investigate the influence of artery geometry parameter on stent fatigue life. The stents’ fatigue resistance ability was calculated using Goodman diagram, and the cycle to failure, the fatigue life, and the fatigue safety factor (FSF) were analyzed. Although the peak top of the von Mises stress was located at the bending area of crowns, the stress distributions were different in the two models. Considering the safety and accuracy, it is necessary to use a realistic geometric model to calculate the stent fatigue performance.

In coronary artery bypass grafting (CABG), competitive flow usually happens when the native coronary has a moderate stenosis. But whether the competitive flow could be affected by the length of this moderate stenosis is still a problem. This study constructs a series of CABG models with different stenosis lengths and uses multi-scale modeling method for numerical simulations. The results contain graft flow, the magnitude of competitive flow, wall shear stress (WSS) and oscillating shear index (OSI). According to the results, it is known that the effect of competitive flow will decrease with the increase of stenosis length. But when the length reaches 15mm, the competitive flow will hardly change anymore.

Computational fluid dynamics analyses were performed on three models which have a giant aneurysm with or without a stenosis. The first is a model with an aneurysm (no stenosis and no stent), the second is a model with a preaneurysm stenosis, and the third is a model with an aneurysm implanted with a stent. The increase in pressure in aneurismal sac caused by a 50% stenosis is about 10.3mmHg at peak systole (comparison between the second model and the first model). It must pay attention to the increase of the pressure for the patient which has an aneurysm accompanied by a stenosis when making the treatment plan. Otherwise, it may cause the aneurysm rupture.

A new modeling approach using two different zero-load geometries (diastole and systole) was introduced to properly model active contraction and relaxation for more accurate stress/strain calculations. Ventricle diastole and systole material parameter values were also determined based on in vivo data. Echo-based computational two-layer left ventricle (LV) models using one zero-load geometry (1G) and two zero-load geometries (2G) were constructed. Material parameter values in Mooney–Rivlin models were also adjusted to match echo LV volume data. Effective Young’s moduli (YM) were calculated for ventricle materials for easy comparison. The 2G models may lead to more accurate ventricle stress/strain calculations and material parameter value estimations.

Patient-specific vessel material properties are in general lacking in image-based computational models. Carotid plaque stress and strain conditions with in vivo material and old material models were investigated (8 patients, 16 plaques). Plaque models using patient-specific in vivo vessel material properties showed significant differences from models using old material properties from the literature on stress and strain calculations. These differences demonstrated that models using in vivo material properties could improve the accuracy of stress and strain calculations which could potentially lead to more accurate plaque vulnerability assessment.

A method which can personalize the lumped parameter model of coronary artery and cardiovascular system based on the non-invasive physiological parameters has been developed. The parameters of system were determined by different physiological parameters. The heart module was determined by aortic pressure and heart rate; the systemic circulation module was determined by cardiac output, height and cardio-ankle vascular index (CAVI), while the CAVI was determined by age and aortic pressure; the coronary module was determined by the target waveforms of coronary flow rate predicted from cardiac output. The considerable results proved that this method could be applied to each patient.

Computer simulations of the blood flow through right coronary arteries with two stenoses in the same arterial segment are carried out to investigate the interactions of serial stenoses, especially the effect of the distal stenosis. Various mathematical models are developed by varying the location of the distal stenosis. The numerical results show that the variation of the distal stenosis has significant impact on coronary hemodynamics, such as the pressure drop, flow shifting, wall shear stress and flow separation. Our simulations demonstrate that the distal stenosis has insignificant effect on the disturbed flow pattern in the regions of upstream and across the proximal stenosis. In a curved artery segment with two moderate stenoses of the same size, the distal stenosis causes a larger pressure drop and a more disturbed flow field in the poststenotic region than the proximal stenosis does. A distal stenosis located at a further downstream position causes a larger pressure drop and a stronger reverse flow.

Cardiovascular disease (CVD) is the leading cause of death in the world. Considerable research has been done linking various mechanical risk factors to plaque progression and rupture. However, methods transforming research results to clinical implementation are limited by time-consuming processes. In this study, a 3D thin-wall (TW) model was developed to approximate the 3D fluid–structure interaction (FSI) model to save time for clinical implementations. Results from one patient data (100 TW models) indicated that the relative errors of the average plaque wall stress (APWS) and strain were less than 7%, meanwhile the correlation results of the TW models were similar with that of the FSI models.

Accurate cap thickness quantification is of fundamental importance for vulnerable plaque detection in cardiovascular research. A segmentation method for intracoronary optical coherence tomography (OCT) image based on least squares support vector machine (LS-SVM) was performed to characterize plaque component borders and quantify fibrous cap thickness. Manual segmentation of OCT images were performed by experts based on combination of virtual-histology intravascular ultrasound (VH-IVUS) and OCT images and used as gold standard. The segmentation methods based on LS-SVM provided accurate plaque cap thickness (an 8.6% error by LS-SVM vs. 71% error by IVUS50) serving as solid basis for plaque modeling and assessment.

Finite element (FE) analysis has been widely used to investigate bone responses to mechanical loading. Research in long bones has been straight forward because modeling of these bones requires only two material properties. Such an FE model may provide an adequate approximation of the anatomy for many cases. However, a more detailed model of skull bones is needed to accurately capture its complex structure of multiple bone pieces and the various mineral densities distributed throughout these bone pieces. Unfortunately, FE model development incorporating both complex geometries and anatomically accurate material properties is both computationally and labor intensive. In this study, a method is proposed to automatically segment micro-computed tomography ($\mu$-CT) scan images of bone pieces to build an FE model of a full swine hemi-skull. Using the Digital Imaging and Communications in Medicine (DICOM) files from scanned bones, the complete geometry of each bone piece is recreated through seven customized processing algorithms. After assembling the bone pieces to form the skull, experimentally derived Young’s modulus values are correlated to grayscale values to produce a detailed FE model for accurate simulation. This detailed skull model can be used to predict strain/stress patterns in response to various loading regimes to facilitate research questions in fracture healing and growth, as well as bone tissue engineering and bone mineral density loss (e.g., osteoporosis).

In multi-lens microscopic cone-beam CT system, different resolutions can be achieved by tuning different objective lens with different magnifications, following the scan field of view (FOV) changed. By this multi-FOV/resolution acquisition strategy, many applications like truncation problem can be solved successfully. However, in order to achieve high-resolution image, the microscopic cone-beam CT system should comply with its geometrical restrictions accordingly, which are not only for that of conventional cone-beam system but also for that of multiple objective lens. Hence, compared with the existed cone-beam system calibration, a specific united calibration method is proposed in this paper. A set of geometrical parameters is presented, which can fully describe whole geometry of conventional cone-beam system as well as multi-lens. By taking use of projective data of circular trajectory of one single ball phantom, a cost function with the described arguments based on the projective lines of the phantom is built up. After solving the cost function, the calibration parameters can be determined optimally. The simulated and experimental results demonstrate that the proposed method shows significant improvements in robustness, precision and implementation.

In order to realize the image fusion and improve the quality of the integration, with the aid of dual modality contrast agent (DMCA), a novel fusion method between ultrasound imaging (US) and magnetic resonance imaging (MRI) was put forward in this paper. Due to US’s strong speckle noise, it was an enormous challenge to fuse US with any other modality images. Under the circumstances, DMCA was used in both US and MRI to strengthen the most important information region of interest. Then, because of Rayleigh distribution of US, a self-adaptive weighted non-negative matrix factorization (SWNMF) scheme is constructed to complete the fusion process. In view of the above-mentioned method, the multiple group comparison tests indicated that US-MRI fusion may be a remarkable method for gaining high-quality fusion image.

In order to investigate the aqueous humor flow based on the coupling effects of fluid–solid–heat under different ocular pressure, four models of rabbit eye were constructed and multi-field coupling vs. fluid–solid interaction was compared in these models. ADINA software was applied to mimic the aqueous humor flow. The contour of temperature demonstrates multiple peaks with larger value near the pupil. The flow pattern is influenced mainly by the variation of ocular pressure. It is necessary to perform multi-field coupling simulation based on dynamic pressure measurement when studying the aqueous humor flow in eyes.

In order to establish a treatment plan to predict therapeutic parameters accurately by radio frequency ablation (RFA) in spinal tumor, we use finite element method (FEM) to build a database in different therapeutic parameters (voltages 10–30V and heating-up time 120–1200s) and structures of electrode (single and multi). For predicting parameters through database, we compare three modes. We acquired 383 sets of data as database. The results show support vector machine (SVM) model is the best one to be applied in treatment planning, in which error rates (0 and 0.33%) were the smallest and correlation coefficients (1.000 and 0.999) were the closest to 1.

The relationship between simultaneous limbs blood pressure differences and Ankle–Brachial Index (ABI) is controversial. This paper aims to investigate the association of limbs blood pressures differences with ABI as the current non-invasive diagnosis method in clinical primary care. A cross-sectional study was performed to analysis the relationship between them. The results showed that $\mathrm{\text{ABI}}<0.9$ was independently associated with inter-arm (OR, 15.469; CI, (1.776–134.773); $P=0.013$) and inter-ankle (OR, 7.189; CI, (1.010–51.179); $P=0.049$) when the systolic blood pressure difference $\ge 15$mmHg. Therefore, the simultaneous measurement of four limbs blood pressure differences can provide an aid for the non-invasive detection method of PAD in clinical primary care.