(Poster 13) Feasibility of Tracking Pelvic Structures in Continent and Incontinent Women

Objectives: Several pelvic floor structures work together as a complex coordinated system to maintain pelvic organ support, sexual function, and urinary continence. In this study, our objectives were to 1) determine the feasibility of using our semi-automated tracking software to track the direction and amount of motion of seven pelvic floor structures between continent and incontinent females, and 2) establish interrater reliability of our semi-automated tracking software using two-dimensional transperineal ultrasound imaging (2D-TPUS).

Methods: A secondary analysis was performed using previously collected 2D-TPUS scans of thirteen female subjects, six continent and seven incontinent (stress, urge, or both). Two raters independently identified seven anatomical regions of interest (ROI): (a) bulbospongiosus, (b) vesical neck, (c) mid urethra, (d) distal urethra, (e) vagina, (f) anorectal angle, (g) external anal sphincter (Figure 1). Movement of these seven ROIs was tracked relative to a reference ROI (pubis) using custom image analysis software during three physiological activities: (1) Valsalva, (2) Cough, and (3) Kegel. Interrater reliability was performed using all videos.

Results: Demographics were similar between continent and incontinent women (age 38.2 ± 16.1 years, BMI 26.0 ± 5.6 kg/m2). All continent subjects were nulliparous, two of the incontinent subjects had vaginal deliveries (four and two, respectively). Our software was able to track the direction and amount of movement of all seven structures during performance of all three physiological activities for all participants (Table 1). Interrater reliability demonstrated that each respective ROI was placed within a mean distance of 0.65 ± 0.25 cm (range 0.24-1.38 cm) between raters. The magnitude of the movement of the mid urethra during a cough differed between continent and incontinent individuals (0.29±0.10 vs 0.46±0.12 cm, p=0.04).

Conclusions: Our semi-automated software was capable of tracking and quantifying motion of all seven pelvic floor structures in both continent and incontinent women during performance of three physiologic tasks. Differences observed in this small sample occurred in the expected direction. This technique can be used in future studies to explore motor control of the pelvic floor.

Figure 1. (a) Colored rectangles indicate seven regions of interest (ROIs). Pubic bone (PB) serves as the reference point and interior dots as features that are tracked. (b) Motion tracking lines at maximal exertion of the Valsalva maneuver based on the upper left corner of the rectangle. ARA, anorectal angle; BSM, bulbospongiosus muscle; EAS, external anal sphincter; U, urethra (d, distal; m, middle; p, proximal); V, vagina.