Nuchal translucency thickness measurement: repeatability using a virtual ultrasound scanner.

Abstract

To use a PC-based virtual ultrasound scanner (VirUS) in the investigation of inter- and intraoperator nuchal translucency (NT) thickness measurement repeatability of experienced ultrasound operators. Realistic fetal ultrasound images of defined NT thickness were simulated with VirUS with emulation of scanner gain and time-gain compensation and gain-dependent echo size changes. A set of 50 images was generated with uniformly distributed NT thickness (range, 1-5 mm at 1-mm intervals) and translucency angle (mean +/- standard deviation of +/- 2.52 degrees +/- 1.85 degrees about the horizontal). Operators (n = 13) measured NT thickness in the image set on three occasions separated by at least 1 day, giving 150 measurements per operator (total measurements, 1950). Inter- and intraoperator repeatabilities were +/- 0.41 mm and +/- 0.22 mm, respectively (at the 95% confidence level). There were significant correlations between repeatability and mean measured NT thickness (r = - 0.72, P = 0.005 at 4-mm interval), between gain and mean measured NT thickness (P </= 0.002, n = 8/13) and between gain and repeatability coefficient (P < 0.01, n = 6/13). VirUS provides a consistent NT audit environment and demonstrates the need to both optimize repeatability vs. mean measured thickness and to set gain consistently. The technique has potential in operator training.