Ultrasound-based personalized hemodynamic modeling of large regions of the peripheral artery using a novel optical tracking approach.

Abstract

Peripheral arterial disease (PAD) affects over 200 million people globally and is commonly diagnosed using Duplex ultrasound (DUS), which is limited to 2D imaging and prone to measurement errors. While CT and MRI provide 3D insights, they are costly and less accessible. Accurate, subject-specific models are essential for reliable hemodynamic assessment using Computational Fluid Dynamics (CFD). We developed a framework combining freehand 2D US with a Polaris Vega ST optical tracking system to reconstruct 3D geometries of the superficial femoral artery (SFA). Phantom validation of the reconstructed geometries showed diameter errors within ±0.1 mm. Ten healthy volunteers and one PAD patient were scanned using a Philips Epiq 7G system. In vivo scan lengths ranged from 54-160 mm, and segmented lumen diameters aligned with previous studies. A semi-automated segmentation pipeline in MATLAB was used to reconstruct the vessel geometries. CFD simulations were performed, and realistic eccentric stenoses were added to the healthy geometries to quantify hemodynamic parameters in stenotic arteries, as typically observed in patients with PAD. CFD models revealed increased pressure drops and flow reductions with higher stenosis severity. Regions prone to atherosclerotic development (atheroprone zones) were identified, which increased in size and severity with stenosis degree. The proposed optical tracking based US framework enables accurate, radiation-free 3D reconstruction of the SFA with longer scan lengths compared to electromagnetic-tracking-based setups, and the resulting CFD models capture key hemodynamic changes and atheroprone regions, highlighting the method's potential for personalized PAD assessment and treatment planning.