Volumetric (3D) imaging reduces inter- and intraobserver variation of fetal biometry measurements
Volumetric (3D) imaging reduces inter- and intraobserver variation of fetal biometry measurements
复制标题
DOI：
10.1002/uog.6321
复制
发表时间：
2009-04-01
影响因子：
7.1
通讯作者：
Lau, T. K.
中科院分区：
文献类型：
--
作者：
Chan, L. W.;Fung, T. Y.;Lau, T. K.
目的比较利用二维（2D）和三维（3D）超声成像的胎儿生物特征测量在观察者间和观察者内的变化。方法本前瞻性研究采用三对接受超声检查培训的医生对中期单胎妊娠进行评估。每位医生使用2D成像分别测量双顶骨直径（BPD）、头围（HC）、腹围（AC）和股骨长度（FL），然后再次使用3D体积数据集。每组医生对12名患者进行评估。观察者间和观察者内变异分别计算为医生对和个体医生进行的成对测量之间差异的标准差。Bland-Altman图用于通过2D和3D方法直观地比较测量偏差和一致性。结果二维和三维超声在观察者间和观察者内的变化较小。与二维超声相比，三维超声下HC、AC和FL的观察内变化明显降低。2D和3D超声测量的观察者间变异无显著差异，但FL在3D超声测量时较低。Bland-Altman图显示，在95%的测量中，2D和3D超声技术对BPD、HC、AC和FL的百分比差异分别在5.3%、4.6%、9.6%和9.6%之内。结论二维和三维胎儿生物特征测量具有可重复性。3D超声的使用显著减少了HC， AC和FL的观察者内部变异，并减少了FL的观察者之间的变异。版权(C) 2009 ISUOG，由John Wiley & Sons， Ltd出版。
Objective TO compare the inter- and intraobserver variation of fetal biometric measurements utilizing two-dimensional (2D) and three-dimensional (3D) ultrasound imaging.Methods This prospective study, utilizing three pairs of doctors trained in sonography, evaluated singleton pregnancies in the mid-trimester. Measurements of the biparietal diameter (BPD), bead circumference (HC), abdominal circumference (AC) and femur length (FL) were taken in duplicate by each doctor using 2D imaging and then again using 3D volume datasets. Each set of paired doctors evaluated 12 patients. Inter- and intraobserver variation were calculated as the SD of the difference between paired measurements performed by the doctor pairs and by the individual doctors, respectively. Bland-Altman plots were used to visually compare measurement bias and agreement by 2D and 3D methods.Results Inter- and intraobserver variation for 2D and 3D ultrasound were small. The intraobserver variation of HC, AC and FL was significantly lower for 3D compared with 2D ultrasound. Interobserver variation was not significantly different when measured with 2D and 3D ultrasound, with The exception of FL, which was lower when measured with 3D ultrasound. The Bland-Altman plots showed that in 95% of the measurements, the percentage difference between 2D and 3D ultrasound techniques was within 5.3%, 4.6%, 9.6% and 9.6% for BPD, HC, AC and FL, respectively.Conclusions 2D and 3D fetal biometric measurements are reproducible. The use of 3D ultrasound significantly reduces intraobserver variation for HC, AC and FL and reduces interobserver variation for FL. Copyright (C) 2009 ISUOG, Published by John Wiley & Sons, Ltd.