Comparison of translabial three-dimensional ultrasonography and magnetic resonance imaging for the grading of levator ani defects

Levator ani defect (LAD) closely correlates with pelvic organ prolapse. This study aimed to compare the LAD grading between 3-dimensional ultrasonography (3D-US) and magnetic resonance imaging (MRI) and investigate the reasons for the difference using 3-dimensional pelvic models. Seventy-two Chinese women who were to undergo repair surgery were assessed by the prolapse staging, 3D-US and MRI. LAD was graded according to the grading systems described with regard to 3D-US (Dietz et al.) and MRI (Delancey et al.) The puborectalis attachment width and the puborectalis thickness were measured on the reconstructed pelvic models offline within the software. The results were analyzed using the weighted kappa and the ANOVA test. The grading systems used for 3D-US and MRI showed the good agreement (kappa = 0.75), whereas the consensus of the extent (ie, partial or complete) of tears showed the moderate agreement (kappa = 0.56). Additionally, iliococcygeus tears detected by MRI (n = 3) accompanied with complete puborectalis tears on the same side. The averaged width of intact puborectalis attachment was 13.75 +/- 3.43 mm. The width of intact puborectalis attachment was remarkably higher than that of the injured attachment (P = .005). The averaged puborectalis thickness was 9.85 +/- 2.13 mm. Comparison of 3D-US and MRI showed the good agreement on LAD grading. The moderate agreement in assessing partial or complete tears resulted from the grading criteria of 3D-US. The morphological characteristics of puborectalis assisted in identifying complete tears.