Abstract WP163: From Static Imaging to Hemodynamic Modelling: Automated Nascet Grading Using Computational Fluid Dynamics

Abstract

Aim: Duplex ultrasound is used for non-invasive assessment of internal carotid artery (ICA) disease and stenosis grading based on The North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria. Computational fluid dynamics (CFD) can be used to model blood flow based on static scans (like CT or MR angiograms), but such models are rarely validated in real clinical settings. We evaluated the accuracy of an automated vascular analysis software that simulates blood flow given a CT or MR angiogram (See-Mode Technologies, Singapore) against NASCET grading done by duplex ultrasound. Methods: 20 carotid arteries from patients with ICA disease were retrospectively analyzed using an automated vascular analysis software according to a blinded protocol. Three-dimensional models of the carotid arteries were reconstructed from MR or CT angiogram DICOM images, and CFD was used for patient-specific modelling of blood flow in the vasculature. Blood flow velocities, specifically ICA peak systolic velocity (PSV) and PSV ratio between the ICA and the common carotid artery (CCA) were obtained from the software and used for stenosis grading based on the NASCET criteria. Bland-Altman analysis was used to assess the agreement between the software and duplex ultrasound. Results: The blood flow simulations were in excellent agreement with duplex ultrasound for both metrics, i.e. ICA PSV (r 2 = 0.936, P < 0.001) and ICA/CCA PSV ratio (r 2 = 0.932, P < 0.001). There was no significant over or under estimation of ICA PSV, but there was a slight over estimation of the ICA/CCA PSV ratio with the software (ICA PSV: bias = 5 cm/s, lower limit of agreement (LOA) = -43 cm/s, upper LOA = 52 cm/s; ICA/CCA PSV ratio: bias = 0.16, lower LOA = -0.54, upper LOA = 0.86). All carotid arteries were correctly classified with the software compared against the NASCET grading done with duplex ultrasound. Conclusion: While CFD is commonly used for modelling blood flow, its accuracy is rarely assessed in real clinical settings. This is the first validated comparison of carotid artery stenosis grading using blood flow simulations against duplex ultrasound. When the data was categorized according to NASCET criteria, all cases were correctly classified with the software compared against duplex ultrasound.