An innovative virtual reality technique for automated human embryonic volume measurements

Abstract

The recent introduction of virtual reality (VR) enables us to use all three dimensions in a three-dimensional (3D) image. The aim of this prospective study was to evaluate an innovative VR technique for automated 3D volume measurements of the human embryo and yolk sac in first trimester pregnancies. We analysed 180 3D first trimester ultrasound scans of 42 pregnancies. Scans were transferred to an I-Space VR system and visualized as 3D 'holograms' with the V-Scope volume-rendering software. A semi-automatic segmentation algorithm was used to calculate the volumes. The logarithmically transformed outcomes were analysed using repeated measurements ANOVA. Interobserver and intraobserver agreement was established by calculating intraclass correlation coefficients (ICCs). Eighty-eight embryonic volumes (EVs) and 118 yolk sac volumes (YSVs) were selected and measured between 5(+5) and 12(+6) weeks of gestational age (GA). EV ranged from 14 to 29 877 mm(3) and YSV ranged from 33 to 424 mm(3). ANOVA calculations showed that when the crown-rump length (CRL) doubles, the mean EV increases 6.5-fold and when the GA doubles, the mean EV increases 500-fold (P < 0.001). Furthermore, it was found that a doubling in GA results in a 3.8-fold increase of the YSV and when the CRL doubles, the YSV increases 1.5-fold (P < 0.001). Interobserver and intraobserver agreement were both excellent with ICCs of 0.99. We measured the human EV and YSV in early pregnancy using a VR system. This innovative technique allows us to obtain unique information about the size of the embryo using all dimensions, which may be used to differentiate between normal and abnormal human development.