Three-dimensional multiplanar transvaginal ultrasound of the cervix in pregnancy.

Abstract

To develop a standard technique for using three-dimensional ultrasound (3D US) to study and evaluate the cervix in pregnant women at high risk for premature delivery, comparing the findings on 3D US with those on conventional two-dimensional ultrasound (2D US). Twenty-one pregnant women at high risk for premature delivery had a total of 37 transvaginal 2D and 3D US examinations of the cervix between 11 and 32 weeks' gestation. A 3D US vaginal probe (5.0-8.0 MHz, Voluson 530D, Medison, Pleasanton, CA, USA) was used. Measurements made from the 2D and 3D US were compared. Of the 37 cervical length measurements in 3D US sagittal plane, seven were shorter and three were longer than on 2D US (varying by 5-15 mm) indicating that the true mid-sagittal plane was not obtained in ten (27%) of the 37 2D US examinations. Of 21 examinations showing funneling, funneling was seen on both 2D and 3D US in 15, but was seen only on 3D US in six. There was a significant (P < 0.05) difference between funnel width as measured in the coronal 3D plane versus 2D US, as well as between funnel width in the coronal 3D US plane versus sagittal 3D US plane (P < 0.05). The cerclage was seen in its entirety in nine of 10 examinations; the 3D US axial plane was most valuable for imaging the cerclage. 3D US appears to offer a more complete assessment of the cervix than 2D US. Multiplanar correlation shows that the standard 2D US sagittal view may under- or over-estimate cervical length. Our preliminary data suggest that 3D US has the potential to improve our understanding of cervical morphology.