Transperineal ultrasound shear wave elastography (SWE) was recently reported as an innovative tool to measure the levator ani muscle (LAM) elastic properties [1]. Our main objective was to assess the intraoperator intersession reproducibility of ultrasound to measure the LAM elastic properties in women. The secondary objectives were to investigate the intrasession reproducibility and to compare intersession reproducibility when considering the mean of three consecutives measures versus one single measure.

We conducted a monocentric prospective study including non-pregnant, nulliparous women. Women with personal history of pelvic floor disorders, body mass index higher than 35Kg.m-2, muscular disease were excluded. Two visits were planned with a time interval ranged from 12h to 7 days. During each visit it was performed a measure of the elastic properties of the right LAM by a single investigator which is a senior Urogynecologist. These measurements were done during three conditions: rest, Valsalva maneuver and perineal contraction. For each measure, the participants lay down in the lithotomy position with an empty bladder. We used an Aixplorer ® device (V12, Supersonic Imagine, France) with a linear SL 18-5 probe (5-18 MHz). The pubic insertion of the right LAM was identified in B-mode with a transperineal approach and then we proceeded to perform the SWE acquisition. Before any measurement, the volunteer performed 2 Valsalva maneuvers with biofeedback to prevent LAM coactivation. For assessments at rest, the women had to relax as much as possible. For assessment during Valsalva, she had to perform a maximal Valsalva maneuver for at least 5s and for assessments at contraction she had to contract her perineum as if she wanted to avoid gas leakage for at least 5s. Data were collected as 5 seconds video clips with three consecutives video clips for each condition: rest, Valsalva and contraction [1,2]. Within each video clip the region of interest was identified and contoured using Matlab scripts (The Mathworks, Inc., 2016) within we measured the shear modulus (SM), in kPa. For measurements at rest and Valsalva maneuver, the mean SM during the video clip was considered whereas for measurement during contraction we considered the maximal SM modulus. For the primary objective, we analyzed the intersession reproducibility for each condition with the Intraclass Correlation Coefficient (ICC), the Standard Error of Measurement (SEM), and the coefficient of variation (CV) serving as main judgement criteria. For this analysis we considered the mean of the three consecutives measurements for each condition. To address our secondary objectives, we investigated the intrasession reproducibility within three consecutives measures using the same methods than for the primary objective. Last, we compared the reproducibility performances when considering the mean of the three consecutive measures or the first f the three measures. Statistical analysis was performed using STATA software (V14IC; Stata Corporation, College Station, TX, USA).  Continuous variables were reported as means and standard deviation (SD). This study was approved by an ethical committee and each woman gave her written free consent before any investigation.

Twenty women were included in this study with a mean age of 23 years (SD=4), a mean body mass index of 22.6 Kg.m-2 (SD= 3.2). the mean interval between 2 visits was 46.6h (SD= 39.6). 
Results about the intersession reproducibility analysis are reported in Table 1 with an excellent reproducibility for assessment at rest and Valsalva maneuver whatever if we considered the mean of the 3 measures or just the first of the three consecutives measures. Reproducibility was fair for measures during contraction. Results about the intrasession reproducibility within the three consecutives measures are reported in Table 2 with an excellent reproducibility at rest and Valsalva maneuver whereas it was moderate during contraction.

The reliability was excellent at rest and Valsalva maneuver when considering the mean of the three consecutives measures or the first one. This result would have been the same if considering the second or the third measure because the intrasession reproducibility was also excellent. Based on this observation, it appears safe to perform one single measure when the objective is to assess the elastic properties for a specific and unique time. If the objective is to investigate changes across the time, it is safer to perform three measures and consider their mean regarding that this seems increase the reliability. The possibility of safely measuring the elastic properties of women’s pelvic floor muscles in a non-invasive way is very innovative and allows some very interesting prospects. As an example, this technology is currently used in a prospective study in a pregnant women cohort and it is expected from this study to provide the first human data about changes in the elastic properties of the pelvic floor during pregnancy and their interactions with the risk of perineal trauma at childbirth [2, 3]. Likely, this technology offers the possibility to measure the elastic properties of the pelvic floor muscles in women suffering from pelvic floor disorders with the prospect of an individual discussion about the treatment strategy based on the patient’s intrinsic characteristics.

Transperineal ultrasound SWE is a reliable tool to investigate the LAM elastic properties in women at rest and Valsalva maneuver but we failed to perform reliable measurement during perineal contraction.

Gachon B et al. In vivo assessment of the levator ani muscles using shear wave elastography: a feasibility study in women. Int Urogynecol J. 2019;30:1179-1186
Gachon B et al. In vivo assessment of the elastic properties of women's pelvic floor during pregnancy using shear wave eastography: design and protocol of the ELASTOPELV study. BMC Musculoskeletal Disord. 2020:21:305.
Gachon B et al. Tissue biomechanical behavior should be considered in the risk of perineal trauma at childbirth. Arch Gynecol Obstet. 2019;300:1821-1826.