Deep Learning-Assisted Two-Dimensional Transperineal Ultrasound for Analyzing Bladder Neck Motion in Women With Stress Urinary Incontinence.

BACKGROUND: No universally recognized transperineal ultrasound parameters are available for evaluating stress urinary incontinence. The information captured by commonly used perineal ultrasound parameters is limited and insufficient for a comprehensive assessment of stress urinary incontinence. Although bladder neck motion plays a major role in stress urinary incontinence, objective and visual methods to evaluate its impact on stress urinary incontinence remain lacking. OBJECTIVE: To use a deep learning-based system to evaluate bladder neck motion using two-dimensional transperineal ultrasound videos, exploring motion parameters for diagnosing and evaluating stress urinary incontinence. We hypothesized that bladder neck motion parameters are associated with stress urinary incontinence and are useful for stress urinary incontinence diagnosis and evaluation. STUDY DESIGN: This retrospective study including 217 women involved the following parameters: maximum and average speeds of bladder neck descent, ß angle, urethral rotation angle, and duration of the Valsalva maneuver. The fitted curves were derived to visualize bladder neck motion trajectories. Comparative analyses were conducted to assess these parameters between stress urinary incontinence and control groups. Logistic regression and receiver operating characteristic curve analyses were employed to evaluate the diagnostic performance of each motion parameter and their combinations for stress urinary incontinence. RESULTS: Overall, 173 women were enrolled in this study (82, stress urinary incontinence group; 91, control group). No significant differences were observed in the maximum and average speeds of bladder neck descent and in the speed variance of bladder neck descent. The maximum and average speed of the ß and urethral rotation angles were faster in the stress urinary incontinence group than in the control group (151.2 vs 109.0 mm/s, P=0.001; 6.0 vs 3.1 mm/s, P <0.001; 105.5 vs 69.6 mm/s, P <0.001; 10.1 vs 7.9 mm/s, P=0.011, respectively). The speed variance of the ß and urethral rotation angles were higher in the stress urinary incontinence group (844.8 vs 336.4, P <0.001; 347.6 vs 131.1, P <0.001, respectively). The combination of the average speed of the ß angle, maximum speed of the urethral rotation angle, and duration of the Valsalva maneuver demonstrated a strong diagnostic performance (area under the curve, 0.87). When 0.481*ß anglea + 0.013*URAm + 0.483*Dval = 7.405, the diagnostic sensitivity was 70% and specificity was 92%, highlighting the significant role of bladder neck motion in stress urinary incontinence, particularly changes in the speed of the ß and urethral rotation angles. CONCLUSIONS: A system utilizing deep learning can describe the motion of the bladder neck in women with stress urinary incontinence during the Valsalva maneuver, making it possible to visualize and quantify bladder neck motion on transperineal ultrasound. The speeds of the ß and urethral rotation angles and duration of the Valsalva maneuver were relatively reliable diagnostic parameters.

Study on the Consistency Between Automatic Measurement Based on Convolutional Neural Network Technology and Manual Visual Evaluation in Intracavitary Ultrasonic Cine of Anterior Pelvic.

OBJECTIVES: This study was to evaluate the application of automatic measurement based on convolutional neural network (CNN) technology in intracavitary ultrasound cine of anterior pelvic. METHODS: A total of 500 patients who underwent pelvic floor ultrasound examination at Peking University Shenzhen Hospital from July 2021 to February 2022 were retrospectively retrieved by the picture archiving and communication system (PACS) system, and 300 cases were used as a training set. The training set was labeled by three experienced ultrasound physicians to train CNN models and develop an automatic measurement software. The remaining 200 cases were used as a test set. Automatic measurement software identified relevant anatomical structures frame by frame and determined the two frames with the greatest difference, calculated the bladder neck descent (BND), urethral rotation angle (URA), and retrovesical angle (RA). Meanwhile, two experienced ultrasound physicians evaluated the resting frame and the maximum Valsalva frame on the cines by manual visual evaluation, labeled the anatomical structures in the corresponding frame, such as the inferoposterior margin of pubic symphysis, the mid-axis of pubic symphysis, bladder contour, and urethra in the front, and calculated BND, URA, and RA. Considering that the residual urine volume (RUV) in the bladder may affect the results, enrolled patients were grouped according to the RUV (10-50 mL, 50-100 mL, and >100 mL). The consistency of the results by automatic measurement and manual visual evaluation was evaluated using the intraclass correlation coefficient (ICC) and the Bland-Altman graph. RESULTS: Of the 200 cases in the test set, 120 cases were successfully identified by the CNN automatic software with a 60% recognition rate. In the case of successful identification, the ICC of manual visual evaluation measurement and automatic measurement was 0.936 (BND), 0.911 (URA), 0.756 (RA in rest), and 0.877 (RA at maximum Valsalva), respectively. In addition, the RUV had a negligible effect on the consistency. The Bland-Altman plot shows the proportion of samples outside the limit was below 5%. CONCLUSIONS: CNN-based automatic measurement software exhibited high reliability in anterior pelvic measurement, which results in a significantly enhanced measurement efficiency.

Fully Automated Localization and Measurement of Levator Hiatus Dimensions Using 3-D Pelvic Floor Ultrasound.

OBJECTIVE: To develop an algorithm for the automated localization and measurement of levator hiatus (LH) dimensions (AI-LH) using 3-D pelvic floor ultrasound. METHODS: The AI-LH included a 3-D plane regression model and a 2-D segmentation model, which first achieved automated localization of the minimal LH dimension plane (C-plane) and measurement of the hiatal area (HA) on maximum Valsalva on the rendered LH images, but not on the C-plane. The dataset included 600 volumetric data. We compared AI-LH with sonographer difference (ASD) as well as the inter-sonographer differences (IESD) in the testing dataset (n = 240). The assessment encompassed the mean absolute error (MAE) for the angle and center point distance of the C-plane, along with the Dice coefficient, MAE, and intra-class correlation coefficient (ICC) for HA, and included the time consumption. RESULTS: The MAE of the C-plane of ASD was 4.81 ± 2.47° with 1.92 ± 1.54 mm. AI-LH achieved a mean Dice coefficient of 0.93 for LH segmentation. The MAE on HA of ASD (1.44 ± 1.12 mm²) was lower than that of IESD (1.63 ± 1.58 mm²). The ICC on HA of ASD (0.964) was higher than that of IESD (0.949). The average time costs of AI-LH and manual measurement were 2.00 ± 0.22 s and 59.60 ± 2.63 s (t = 18.87, p < 0.01), respectively. CONCLUSION: AI-LH is accurate, reliable, and robust in the localization and measurement of LH dimensions, which can shorten the time cost, simplify the operation process, and have good value in clinical applications.