Simultaneous quantitative evaluation of both iliac arteries and veins via accelerated four-dimensional flow in healthy controls and patients with deep vein thrombosis: a pilot study

Background: Understanding iliac hemodynamics in both healthy individuals and patients with vascular diseases is important from both physiological and pathological perspectives, and the four-dimensional (4D) flow technique is a valuable tool for the accurate quantification of hemodynamics. However, conventional 4D flow acquisition is time-consuming. Therefore, this study aimed to evaluate the hemodynamics of the iliac arteries and veins using both conventional single-velocity encoding (VENC) and compressed sensing (CS)-accelerated dual-VENC 4D flow sequences in healthy controls (HCs) and patients with deep vein thrombosis (DVT). Methods: This study included HCs and patients with DVT between December 2024 and April 2025. Two conventional single-VENC 4D flow sequences with high and low VENCs and a dual-VENC CS-accelerated sequence were performed and postprocessed by two experienced readers. Quality control, including test-retest reliability, agreement between techniques, and intra- and inter-reader agreement, was performed, and the correlation between age and flow parameters was also calculated. Results: The acquisition time for conventional high-VENC, low-VENC, and CS dual-VENC 4D flow was 10.6, 10.7, and 8.8 min, respectively. In the quality control analysis, the intraclass correlation coefficients (ICCs) for test-retest reliability, intraobserver, and interobserver agreements were all >0.85 (P≤0.001). In arterial segments, both velocity and wall shear stress (WSS) measured by the CS methods were lower than those obtained by conventional methods, with the underestimation percent ranging from 2.8% to 11.8% in HCs and from 1.9% to 14.8% in patients with DVT. For venous segments, CS-based measurements underestimated velocity and WSS by 1.1-12.0% in HCs and 1.3-14.2% in patients with DVT. Negative correlations were found between age and arterial velocity (R=-0.33), arterial WSS (R=-0.34), and venous velocity (R=-0.29) (all P values <0.05). Conclusions: The study results indicate that the CS dual-VENC 4D flow sequence is feasible for rapid simultaneous assessment of iliac arteries and veins in both HCs and patients with DVT. However, the underestimation caused by CS should be noted.