P06.05: Learning curve for nuchal translucency measurement by semi‐automated and manual methods

To assess the learning curves for nuchal translucency thickness as measured by both manual and semi-automated methods. Fetal nuchal translucency was obtained by an experienced sonologist according to the FMF guidelines in a cohort of 32 consecutive first trimester fetuses at 11–13 weeks, and calculated by both manual and semi-automated methods (GE Voluson E6). Subsequently, two sonographer trainees with no prior experience in first trimester ultrasound examined archived images in separated occasions and measured the nuchal translucency first by the manual method and secondly with the semi-automated method. Each of the trainees was blinded to any pre-existing results. The average difference between the two trainees and the expert in both manual and semi-automated methods was calculated. A difference below 15% was considered as an accurate measurement. The average learning curve was delineated using the cumulative sum analysis (CUSUM). CUSUM plots demonstrated that the mean learning curve was achieved by 19 and 25 measurements for semi-automated and manual methods, respectively. For an inexperienced sonographer to become competent in nuchal translucency measurement in stored images, a minimum number of 19 and 25 measurements was required for the semi-automated and manual method, respectively.