Experimental Validation of Enhanced Magnetic Resonance Imaging (EMRI) Using Particle Image Velocimetry (PIV)

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.