Synthetic Vascular Ultrasound Imaging through Coupled Fluid-Structure Interaction and Ultrasound Simulations

摘要

Although ultrasonic imaging is commonly applied in cardiovascular research and clinical practice, current blood flow and vessel wall imaging methods are still hampered by several limitations. We developed a simulation environment integrating ultrasound (US) and fluid-structure interaction (FSI) simulations, allowing construction of synthetic US-images based on physiologically realistic behavior of an artery. An in-house code was developed to strongly couple the flow solver Fluent and structural solver Abaqus using an Interface Quasi-Newton technique. A distensible tube, representing the common carotid artery (length 5cm, inner diameter 6 mm, thickness 1 mm), was simulated. A mass flow inlet boundary condition, based on flow measured in a healthy subject, was applied. A downstream pressure condition, based on a non-invasively measured pressure waveform, was used. US-simulations were performed with Field II, allowing to model realistic transducers and scan sequences as used in clinical vascular imaging. To this end, scatterers were "seeded" in the fluid and structural domain and propagated during the simulated scan procedure based on flow and structural displacement fields from FSI. Simulations yielded raw ultrasound (RF) data, which were processed for arterial wall distension and shear rate imaging. Our simulations demonstrated that (i) the wall distension application is sensitive to measurement location (highest distension found when tracking the intima-lumen transition); (ii) strong reflections between tissue transitions can potentially cloud a correct measurement; (iii) maximum shear rate was underestimated during the complete cardiac cycle, with largest discrepancy during peak systole; (iv) due to difficulties measuring near-wall velocities with US, shear rate reached its maximal value at a distance from the wall (0.812 mm for anterior and 0.689 mm for posterior side). We conclude that our FSI-US simulation environment provides realistic RF-signals which can be processed into ultrasound-derived medical images and measurements.