A Novel Methodology for Semi-automatic Measurement of Arterial Stiffness by Doppler Ultrasound: Clinical Feasibility and Reproducibility.

The conventional manual approach to measurement of aortic pulse wave velocity (PWV) by Doppler ultrasonography is time consuming and operator dependent. Here we report a new semi-automated methodology for more efficient and objective measurement of aortic PWV and results of tests of its clinical feasibility and reproducibility. Carotid-femoral pulse wave velocity (cfPWV) was measured in 50 patients with suspected coronary artery disease (aged 59.2 ± 10.0 y, 36 males) by three independent observers, including two experienced sonographers and one cardiologist without ultrasonographic experience. The cfPWV measured by the semi-automatic method (cfPWV) was compared with reference values obtained by averaging measurements by two experienced sonographers using the conventional standard manual method (cfPWV). Measurements of cfPWV were feasible in all 50 patients and exhibited excellent agreement with averaged cfPWV from the two experienced sonographers, with an intraclass correlation coefficient (ICC) of 0.915 (95% confidence interval: 0.876-0.942). The inexperienced observer-measured cfPWV did not differ from the cfPWV measured by the two experienced sonographers (8.04 ± 1.29 vs. 8.14 ± 1.32 m/s, p > 0.05), with a high consistency by ICC of 0.877 (0.793-0.928). Bland-Altman plots further illustrated the good agreement between the two methods and good intra- and inter-observer reproducibility. Time consumption for cfPWV measurement using the new method was significantly less than that for the manual method (122 ± 35 s vs. 455 ± 105 s, p < 0.0001), saving about 73% of the time. This new semi-automatic methodology for aortic PWV measurement not only has an accuracy similar to that of the conventional standard manual method but is also highly feasible and time saving. It may provide a reliable, simple and reproducible approach to arterial stiffness evaluation in clinical settings.