Exo-anal imaging of the anal sphincter: a comparison between introital and transperineal image acquisition.

Three-dimensional exoanal ultrasound imaging of the anal sphincter may be obtained transperineally with a convex probe, or at the introitus with a transvaginal probe. We hypothesised that introital acquisition would yield better quality and more reproducible evaluation. We acquired three 3D volumes of the anal sphincter (one transperineal transverse with a 4- to 8-MHz convex probe and two introital with a 5- to 9-MHz probe in transverse and mid-sagittal view) in 20 representative women attending the gynaecology clinic. Each 3D dataset was anonymised and hence blinded for clinical data and for acquisition method. Images were analysed off-line by two expert specifically trained ultrasonographers in a random order to assess image quality, sphincter integrity and sphincteric measurements. We assessed the intra- and interrater agreement by the Cohen's kappa (κ) and by the intraclass correlation coefficient for categorical and continuous variables respectively. The mid-sagittal introital acquisition had most inconclusive images owing to unsatisfactory quality, on which raters agreed (К = 0.80). Subsequently, agreement in the anal sphincter evaluation between transverse introital and transperineal acquisitions was compared. Agreement on internal anal sphincter gap was excellent for both transverse introital (К = 0.87) and transperineal acquisition (К = 0.93). Agreement on external anal sphincter discontinuity was excellent for the transperineal acquisition (К = 0.87) and good for the transverse introital acquisition (К = 0.73). Intra- and interrater agreement of external and internal anal sphincteric measurements were best for transperineal acquisitions. In our hands, transperineal acquisition with a 4- to 8-MHz probe performed better than introital acquisition with a 5- to 9-MHz probe in the assessment of the anal sphincter complex.