A Phantom Experiment to determine the repeatability of two-dimensional (2D) Pulsed Waved Doppler (PWD) indices using automatic and manual measurement.

Abstract

Objectives The purpose of this study was to show the repeatability of PWD-derived indices in a wall-less, single-vessel flow phantom experiment, using two different methods of PWD measurement: automatic and manual mode. Methods The phantom was constructed in a Perspex box containing a single channel which was initially created using a silicone tube (5 mm and 6 mm, inner and outer diameter respectively). The tubing was positioned centrally within tissue-mimicking material (TMM) inclined at 45o, then removed once the TMM was set. A 3-roller peristaltic pump was used to circulate blood-mimicking fluid at 45 revolutions/min representing a heart rate (HR) of 135 B.M. 4 acquisitions of 18 cardiac cycles were stored using a GE Voluson E8 with a C1-5-D transducer. Each cardiac cycle was measured individually using automatic and manual mode; 72 waveforms in triplicate for one observer and 18 in triplicate by two experienced observers for intra-observer and inter-observer repeatability, respectively. Measurement differences were assessed by Bland and Altman plots and intra-class correlation coefficients (ICC). Differences between techniques of systolic/diastolic (S/D) ratio, pulsatility index (PI) and resistance index (RI) were analysed by paired t-test. Results Mean S/D, PI, and RI (3.33, 1.02 and 0.70) measured by automatic mode were identical for both observers, whereas by manual mode all indices were significantly different (PI; p = 0.024; p Conclusions Variability within PWD parameters was found depending on the annotation mode used. Using the automatic method lower values, better concordance, less variation and more repeatable results were demonstrated, despite a lack of perfection observed. Interpretation of indices derived from manually annotated waveforms should be interpreted with caution.