Search

Flow-sensitive four-dimensional Cardiovascular Magnetic Resonance Imaging (4D Flow CMR) has increasingly been utilised to characterise patients' blood flow, in association with patiens' state of health and disease, even though spatial and temporal resolutions still constitute a limit. Computational fluid dynamics (CFD) is a powerful tool that could expand these information and, if integrated with experimentally-obtained velocity fields, would enable to derive a large variety of the flow descriptors of interest. However, the accuracy of the flow parameters is highly influenced by the quality of the input data such as the anatomical model and boundary conditions typically derived from medical images including 4D Flow CMR. We previously proposed a novel approach in which 4D Flow CMR and CFD velocity fields are integrated to obtain an Enhanced 4D Flow CMR (EMRI), allowing to overcome the spatial-resolution limitation of 4D Flow CMR, and enable an accurate quantification of flow. In this paper, the proposed approach is validated in a U bend channel, an idealised model of the human aortic arch. The flow patterns were studied with 4D Flow CMR, CFD and EMRI, and compared with high resolution 2D PIV experiments obtained in pulsatile conditions. The main strengths and limitations of 4D Flow CMR and CFD were illustrated by exploiting the accuracy of PIV by comparing against PIV velocity fields. EMRI flow patterns showed a better qualitative and quantitative agreement with PIV results than the other techniques. EMRI enables to overcome the experimental limitations of MRI-based velocity measurements and the modelling simplifications of CFD, allowing an accurate prediction of complex flow patterns observed experimentally, while satisfying mass and momentum balance equations.

The analysis of the blood flow in the great thoracic arteries does provide valuable information about the cardiac function and can diagnose the potential development of vascular diseases. Flow-sensitive four-dimensional flow cardiovascular magnetic resonance imaging (4D flow CMR) is often used to characterize patients' blood flow in the clinical environment. Nevertheless, limited spatial and temporal resolution hinders a detailed assessment of the hemodynamics. Computational fluid dynamics (CFD) could expand this information and, integrated with experimental velocity field, enable to derive the pressure maps. However, the limited resolution of the 4D flow CMR and the simplifications of CFD modeling compromise the accuracy of the computed flow parameters. In this article, a novel approach is proposed, where 4D flow CMR and CFD velocity fields are integrated synergistically to obtain an enhanced MR imaging (EMRI). The approach was first tested on a two-dimensional (2D) portion of a pipe, to understand the behavior of the parameters of the model in this novel framework, and afterwards in vivo, to apply it to the analysis of blood flow in a patient-specific human aorta. The outcomes of EMRI are assessed by comparing the computed velocities with the experimental one. The results demonstrate that EMRI preserves flow structures while correcting for experimental noise. Therefore, it can provide better insights into the hemodynamics of cardiovascular problems, overcoming the limitations of MRI and CFD, even when considering a small region of interest. EMRI confirmed its potential to provide more accurate noninvasive estimation of major cardiovascular risk predictors (e.g., flow patterns, endothelial shear stress) and become a novel diagnostic tool.

Validation of computational models using in vitro phantoms is a nontrivial task, especially in the replication of the mechanical properties of the vessel walls, which varies with age and pathophysiological state. In this paper, we present a novel aortic phantom reconstructed from patient-specific data with variable wall compliance that can be tuned without recreating the phantom. The three-dimensional (3D) geometry of an aortic arch was retrieved from a computed tomography angiography scan. A rubber-like silicone phantom was manufactured and connected to a compliance chamber in order to tune its compliance. A lumped resistance was also coupled with the system. The compliance of the aortic arch model was validated using the Young's modulus and characterized further with respect to clinically relevant indicators. The silicone model demonstrates that compliance can be finely tuned with this system under pulsatile flow conditions. The phantom replicated values of compliance in the physiological range. Both, the pressure curves and the asymmetrical behavior of the expansion, are in agreement with the literature. This novel design approach allows obtaining for the first time a phantom with tunable compliance. Vascular phantoms designed and developed with the methodology proposed in this paper have high potential to be used in diverse conditions. Applications include training of physicians, pre-operative trials for complex interventions, testing of medical devices for cardiovascular diseases (CVDs), and comparative Magnetic-resonance-imaging (MRI)-based computational studies.