Kalayeh Kourosh, Fowlkes J Brian, Sack Bryan S, LaCross Jennifer, Daignault-Newton Stephanie, Schmidt Payton, Tai Haowei, Schultz William W, Ashton-Miller James A, DeLancey John O
Department of Radiology, University of Michigan, Ann Arbor, Michigan, USA.
Department of Urology, University of Michigan, Ann Arbor, Michigan, USA.
J Ultrasound Med. 2025 Jul;44(7):1213-1227. doi: 10.1002/jum.16676. Epub 2025 Mar 17.
Stress urinary incontinence (SUI) is a prevalent condition that can significantly affect quality of life. Urethral mobility is an important factor in SUI and transperineal ultrasound (TPUS) imaging can provide clear visualization of this movement; however, its quantification has been limited. An automated system to track and quantify urethral movement could provide richer information and reduce inter-observer effects on measurements.
As proof-of-concept for technique development, we used TPUS cine loops obtained on commercial scanners (GE Healthcare and Philips Healthcare) from consented research volunteers. We developed the tracking software based on fundamental concepts from computer vision, specifically corner detection and optical flow-based tracking algorithms. In doing so we account for inadvertent probe movements by using the symphysis pubis as a reference coordinate system.
The system successfully tracks the motion of the urethra during Valsalva maneuvers. It accurately captures and quantifies complex movements, including directional shifts, rotations, displacement vectors of different structures, and the trajectory of motion. These measurements are corrected for any probe movement. We demonstrated the system's efficiency and reliability in near real-time analysis across various ultrasound platforms and video formats. The intraclass correlation coefficients exceeded 0.89 and 0.5 for intra- and inter-rater reliability, respectively.
By providing detailed, objective measurements of urogenital movement, this approach has potential to advance the understanding, diagnosis and treatment of SUI, which in turn, can help tailor more effective treatment strategies. This methodology paper confirms the feasibility of automated quantification of urethral mobility.
压力性尿失禁（SUI）是一种常见病症，会显著影响生活质量。尿道活动度是压力性尿失禁的一个重要因素，经会阴超声（TPUS）成像能够清晰显示这种活动；然而，其量化一直存在局限。一个用于跟踪和量化尿道活动的自动化系统可以提供更丰富的信息，并减少观察者间测量差异的影响。
作为技术开发的概念验证，我们使用了从商业扫描仪（通用电气医疗集团和飞利浦医疗）获取的经同意参与研究的志愿者的经会阴超声动态图像。我们基于计算机视觉的基本概念，特别是角点检测和基于光流的跟踪算法，开发了跟踪软件。在此过程中，我们以耻骨联合作为参考坐标系来校正探头的意外移动。
该系统成功跟踪了瓦尔萨尔瓦动作期间尿道的运动。它能准确捕捉并量化复杂运动，包括方向偏移、旋转、不同结构的位移矢量以及运动轨迹。这些测量值会针对任何探头移动进行校正。我们在各种超声平台和视频格式的近实时分析中展示了该系统的效率和可靠性。组内相关系数在评估者内和评估者间可靠性方面分别超过了0.89和0.5。
通过提供泌尿生殖系统运动的详细、客观测量，这种方法有潜力促进对压力性尿失禁的理解、诊断和治疗，进而有助于制定更有效的治疗策略。这篇方法学论文证实了自动量化尿道活动度的可行性。
