Validation of an automated software (Smartpelvic™) in assessing hiatal area from three dimensional transperineal pelvic volumes of pregnant women: comparison with manual analysis

objectives: the aim of this investigation was to evaluate the agreement between a manual and an automatic technique in assessing levator hiatus area (LHA) during pregnancy from three-dimensional (3D) pelvic floor volumes obtained by trans-perineal ultrasound (TPUS). methods: 3D volumes were acquired during rest, maximum pelvic floor contraction and valsalva maneuver from 66 pregnant women. manual selection of LHA and automatic software (Smart Pelvic™) were applied on TPUS volume starting from a C-plane view. to evaluate intra- and inter-observer variability measurements of LHA were performed twice by the same operator and once by a second sonographer. reference hiatal contours obtained manually by the first operator were compared with the automated ones. reproducibility was evaluated by intraclass correlation coefficients (ICC) and bland-altman plots. results: LHA measurement, using automatic software, achieved excellent intra-observer and inter-observer reproducibility in pregnant women both at rest and after dynamic analysis (ICC>0.9). further, an excellent agreement resulted between manual selection of the LHA and automatic imaging (ICC>0.9). the average time taken to obtain LHA manually was significantly longer when compared to the automatic analysis (p≤0.0001). conclusions: smart pelvic software resulted from a reliable method for automatically measuring the LHA, showing high reproducibility and accuracy.