Reliability and Accuracy of Tomographic 3-D Ultrasound for Grading Vessel Stenosis: A Phantom Study

Abstract

The aim of this phantom study was to assess the accuracy of 3-D tomographic ultrasound (t3DUS) for grading stenosis, using the manufacturer's measurements as the gold standard. The percentage of maximum stenosis was obtained using 2-D ultrasound (2DUS) and t3DUS imaging techniques on a peripheral vascular phantom, including channels with 50%, 75% and 90% stenosis. The inter-observer reproducibility of t3DUS for grading stenosis was assessed using the intraclass correlation coefficient (ICC) and Bland-Altman plots. Mean and mean differences were used to evaluate the accuracy of 2DUS and t3DUS in measuring maximum stenosis in all channels. Inter-operator agreement was excellent, with an ICC value of 0.99 (95% confidence interval: 0.994-0.998, p < 0.001). Bias in measurements was -0.59 ± 2.01% (95% limits of agreement: 4.54, 3.36). The mean difference (MD) between maximum stenosis measurements and reference values for all channels was lower in t3DUS than in 2DUS (t3DUS MD: +1.01%, diameter reduction 2DUS MD: -6.10%; area reduction 2DUS MD: +8.20%). Tomographic 3DUS is a reproducible and accurate imaging method for grading stenosis. The current B-mode 2DUS stenosis grading criteria used in vascular assessment may be underestimating or overestimating the percentage stenosis. Further phantom and human studies investigating the reliability of t3DUS for grading stenosis and other metrics including plaque volume are required.

Keywords: 2-D ultrasound; Tomographic 3-D ultrasound; Vascular phantom; Vessel stenosis.