P19.08: Prospective validation of fetal weight estimation models using fractional limb volume

Abstract

ultrasonography alone. Each biometry was measured twice by each operator. All images were analyzed by two experienced reviewers blinded to the operator’s identity according to the score-based objective evaluation criteria. Results: Fifty consecutive fetuses were evaluated with gestation age of 17 to 34 weeks. The interobserver, intraobserver and intermethod variability for 2D ultrasonography by an experienced operator (2DE), 2D and 3D ultrasonography by an inexperienced operator (2DI and 3DI) was small (all ICCs >= 0.991). Comparing 2DE and 2DI, there were no differences in the image quality scores, but the inexperienced operator took more time to measure the biometry. The largest difference was noted in the measurement of BPD/HC (1 minute and 12 seconds, P < 0.001). A non-significantly higher proportion of fetal biometric measurements by 3DI than 2DI were within 1mm of the measurements by 2DE. There were no differences in the mean image quality scores of fetal biometry between 2DE and 3DI. The mean time (seconds) required to measure BPD, HC, AC and FL was shorter by 3DI than by 2DI (67.2 vs.97.0, 64.6 vs.97.0, 60.1vs.95.1; all P < 0.001), but was significantly longer than by 2DE. The corresponding figures were 24.3, 24.3, 27.9, and 27.2. Conclusion: For an inexperienced operator, fetal biometric measurements using saved 3D volumes can be faster than 2D ultrasonography but with similar accuracy and reproducibility.