Andrzej Polanczyk

A 3D model of thrombus formation in a stent-graft after implantation in the abdominal aorta.

Here we present a 3D kinetic model of thrombus formation in an endovascular prosthesis after implantation in an abdominal aorta with an aneurysm. The computational fluid dynamic technique (CFD) was used to determine the process of thrombus formation and growth in the stent-graft on the basis of the medical data from computed tomography angiography and Doppler ultrasound examination of 10 patients. The Quemada model was used to describe rheological properties of blood. Results of the CFD simulations were validated based on actual data from patients with diagnosed thrombi in aortic implants. The results show that the elaborated CFD model correctly predicted thrombus formation, shape and deposition site in an endovascular prosthesis. The developed CFD model of thrombus growth can be applied to predict the risk of thrombus formation in stent-grafts and assist in selection of geometry of the endovascular prosthesis to reduce possible complications after stent-graft implantation using only basic medical data.

A new approach for the pre-clinical optimization of a spatial configuration of bifurcated endovascular prosthesis placed in abdominal aortic aneurysms

Complexity of the spatial configuration of an aortic implant with bifurcation in the distal part is related to changes in blood hemodynamic in the area of bifurcation which may disturb blood flow and lead to thrombus formation. This study was designed to characterize parameters which define spatial configuration of an aortic implant for which the risk of thrombus formation is the smallest. We used AngioCT data from 74 patients, aged 55 ±10 years, after endovascular procedure to prepare 3D geometries of stent-grafts. Computational Fluid Dynamics (CFD) simulations were used to reconstruct blood hemodynamic and simulate thrombus formation. Next, geometric parameters of stent-grafts included the ratio of volume of upper part to the bifurcations, the relation of inlet and outlet diameters of a stent-graft and deformations in the iliac part of the stent-graft were analyzed. We also analyzed tortuosities (spiral twisting of the flow around the flow direction) and bends (the largest angulation in distal part of a stent-graft). The CFD results were confronted with AngioCT data to verify if computer generated thrombus appeared in particular patient. Additionally, geometric parameters of analyzed stent-grafts were used to propose a mathematical tool for prediction of thrombus appearance. The results showed that tortuosities and bends of a stent-graft had the highest impact on thrombus formation. Formation of thrombi was observed in 22% to 31% of cases (at blood hematocrit Hct = 40%) even for small values of tortuosities and bends indicating that these parameters are dominant in determining blood clotting. Our calculated results overlapped with clinical data in 80% to 91%. Therefore, we conclude that tortuosities and bends have high impact on thrombus formation and should be under special attention during stent-graft recommendation and patients’ follow-ups.

Computational Fluid Dynamics as an Engineering Tool for the Reconstruction of Hemodynamics after Carotid Artery Stenosis Operation: A Case Study

Background and objectives: Brain ischemic stroke is caused by impaired or absolutely blocked blood flow into the brain regions. Despite the large number of possible origins, there is no general strategy for preventive treatment. In this paper, we aimed to predict the hemodynamics in a patient who experienced a critical stenosis operation in the carotid artery. This is a unique study where we used medical data together with the computational fluid (CFD) technique not to plan the surgery, but to predict its outcome. Materials and Methods: AngioCT data and blood perfusion of brain tissue (CT-perfusion) together with CFD technique were applied for stroke formation reconstruction in different clinical conditions. With the use of self-made semiautomatic algorithm for image processing and 3DDoctror software, 3D-vascular geometries before and after surgical intervention were reconstructed. As the paper is focused on the analysis of stroke appearance, apparent stroke was simulated as higher and lower pressure values in the cranial part due to different outcomes of the surgical intervention. This allowed to investigate the influence of spatial configuration and pressure values on blood perfusion in the analyzed circulatory system. Results: Application of CFD simulations for blood flow reconstruction for clinical conditions in the circulatory system accomplished on average 98.5% and 98.7% accuracy for CFD results compared to US-Doppler before and after surgical intervention, respectively. Meanwhile, CFD results compared to CT-perfusion indicated an average 89.7% and 92.8% accuracy before and after surgical intervention, respectively. Thus, the CFD is a reliable approach for predicting the patient hemodynamics, as it was confirmed by postoperative data. Conclusions: Our study indicated that the application of CFD simulations for blood flow reconstruction for clinical conditions in circulatory system reached 98% and 90% accuracy for US-Doppler and CT-perfusion, respectively. Therefore, the proposed method might be used as a tool for reconstruction of specific patients’ hemodynamics after operation of critical stenosis in the carotid artery. However, further studies are necessary to confirm its usefulness in clinical practice.

3D Blood Vessels Reconstruction Based on Segmented CT Data for Further Simulations of Hemodynamic in Human Artery Branches

Abstract We aimed at the reconstruction of the branches of human aortic arch for blood perfusion analysis used later in the Computational Fluid Dynamic (CFD). The reconstruction was performed based on segmentation results obtained from CT data. Two segmentation algorithms, region growing and level set were implemented. Obtained binary segmentation results were next evaluated by the expert and corrected if needed. The final reconstruction was used for preparation of a numerical grid and for further calculation of blood hemodynamic. The collected data composed of blood velocity and blood flow rate in function of time were compared with USG-Doppler data. Results demonstrate that proposed algorithm may be useful for initial reconstruction of human cardiac system, however its accuracy needs to be improved as further manual corrections are still needed.

Evaluating an algorithm for 3D reconstruction of blood vessels for further simulations of hemodynamic in human artery branches

We aimed at verification of a novel algorithm for semiautomatic reconstruction of the branches of human aortic arch for blood perfusion analysis used later in the Computational Fluid Dynamic (CFD). The algorithm works on initially adjusted CT scans divided into segments. Obtained binary segmentation results were next checked manually and corrected if needed. The final reconstruction was used for preparation of a numerical grid and for further calculation of blood hemodynamic. The collected data composed of blood velocity and blood flow rate in function of time were compared with USG-Doppler data. Preliminary results demonstrate that proposed algorithm may be useful for initial reconstruction of human cardiac system, however its accuracy needs to be improved as further manual corrections are still needed.

A novel method for describing biomechanical properties of the aortic wall based on the three-dimensional fluid-structure interaction model

OBJECTIVES Our goal was to present a novel non-invasive approach for assessment of aortic wall displacement to describe its biomechanical properties during the cardiac cycle. METHODS The fluid-structure interaction (FSI) technique was used to reconstruct aortic wall displacement based on computed tomography angiography and 2-dimensional speckle-tracking technique (2DSTT) data collected from 20 patients [10 with healthy aortas (AA) and 10 with abdominal aortic aneurysms (AAAs)]. The mechanical properties of the wall of the aorta were described by the Yeoh hyperelastic materials model with α and β parameters, and wall displacement was determined with 2DSTT. The mechanical parameters of the wall of the aorta in the FSI model were automatically updated in the calculation loop until the calculated and clinically measured wall movements were the same. RESULTS Results showed 98% accuracy of FSI compared to 2DSTT for AA and AAA (P > 0.05). The mean wall deformation for AA was 2.45 ± 0.12 mm and 2.49 ± 0.10 mm for FSI and 2DSTT, respectively (P = 0.40), whereas that for AAA was 2.84 ± 0.44 mm and 2.88 ± 0.45 mm, respectively (P = 0.83). The FSI analysis indicated that the α and β parameters for AA were equal to 14.35 ± 1.30 N⋅cm-2 and 9.33 ± 1.08 N⋅cm-2, respectively; and for AAA, α was 11.00 ± 0.49 N⋅cm-2 and β was 79.46 ± 4.32 N⋅cm-2. CONCLUSIONS The FSI technique may be successfully applied to assess the mechanical parameters of patient-specific aortic walls using computed tomography angiographic and 2DSTT measurements.

3-D CFD simulations of hydrodynamics in the Sulejow dam reservoir

Abstract This paper reports the processes by which a single-phase 3-D CFD model of hydrodynamics in a 17-km-long dam reservoir was developed, verified and tested. A simplified VOF model of flow was elaborated to determine the effect of wind on hydrodynamics in the lake. A hexahedral mesh with over 17 million elements and a k-ω SST turbulence model were defined for single-phase simulations in steady-state conditions. The model was verified on the basis of the extensive flow measurements (StreamPro ADCP, USA). Excellent agreement (average error of less than 10%) between computed and measured velocity profiles was found. The simulation results proved a strong effect of wind on hydrodynamics in the lake, especially on the development of the water circulation pattern in the lacustrine zone.