Abstract:

The prevalence of female pelvic floor dysfunction (PFD) ranges from 17.8% to 74.07%, which severely compromises the life quality and mental health of women. The screening modalities commonly used for PFD, such as pelvic organ prolapse quantification (POP-Q) staging system, standardized questionnaires, two-dimensional ultrasound and static MRI, are only capable of evaluating the static anatomical structure, yet they are highly subjective and lack the dynamic and functional assessment. Although emerging screening techniques exhibit respective advantages, they also present inherent limitations. High-density electromyography features high spatial resolution and enables precise localization of muscle injury, but entails high equipment costs. Three-dimensional/four-dimensional ultrasound and dynamic MRI allow a dynamic and functional evaluation, yet they are constrained by high costs and limited accessibility in primary care settings. Biomarker detection (e.g., 8-hydroxy-2'-deoxyguanosine, etc.) can provide molecular-level diagnostic evidence, but it still requires further validation with large clinical samples. Artificial intelligence prediction models demonstrate promising predictive performance, but lack the standardized multicenter validation. PFD screening is undergoing a transition from single-technique evaluation to multimodal integration. The establishment of a multimodal integration strategy, implementation of a hierarchical screening pathway, and formulation of individualized screening protocols are expected to achieve early detection, precise diagnosis, early intervention of PFD, and ultimately improve the life quality and prognosis of PFD patients.

Key words: Biomarkers, Artificial intelligence, Symptom assessment, Diagnostic screening programs, Pelvic floor dysfunction