Pelvic floor levator hiatus measurements: MRI versus ultrasound

The objective of this study is to compare levator hiatus measurements between pelvic magnetic resonance imaging (MRI) and pelvic ultrasound (US) imaging modalities.

We performed pelvic MRI and 3-dimensional US in 37 asymptomatic nulliparous women. For the MRI protocols, we performed axial and sagittal sequences at rest. We then obtained sagittal sequences during Kegel squeeze and Valsalva maneuvers. Blinded to the findings of the MRI, we obtained 3-dimensional pelvic US images using a perineal approach at rest, Kegel and Valsalva maneuvers. Finally, we measured the levator hiatus in both sagittal and axial planes.

For the resting sagittal measurements, the mean levator hiatus measurement using MRI (5.0 cm; SD, 0.8) is significantly greater than that using US (4.4 cm; SD, 0.6; P < 0.05). Although the absolute mean levator measurements between the 2 modalities are significantly different, this difference is not influenced by the magnitude of the measurements as noted in Bland-Altman plots of the limits of agreement.

We found that the MRI measurements obtained from the sagittal images were consistently greater than the corresponding US images. In contrast, there was not the same consistency of difference between MRI and US for the axial images. This suggests possible variation in acquisition planes for axial images or interpretation of landmarks for the sagittal images.