To propose an automated transperineal ultrasound analysis using deep learning for pelvic floor dysfunction assessment after total hysterectomy, addressing the limitations of current manual methods.
Key Findings:
The model achieved an average Dice coefficient of 88.67 ± 1.96% in segmentation, indicating high accuracy.
Key point localization errors were controlled within 2 mm, demonstrating precision in anatomical landmark detection.
Automatic measurement results showed high consistency with manual annotations, with Pearson correlation coefficients up to 0.92, suggesting reliability in clinical settings.
Interpretation:
The proposed method enables structured, automated, and objective TPUS image analysis, significantly reducing manual intervention.
Limitations:
Validated only in patients with benign diseases, which may limit generalizability.
Transferability to gynecologic oncology patients is suggested but not yet tested, indicating a need for further validation.
Conclusion:
The method provides a reliable tool for postoperative functional monitoring and therapeutic evaluation after total hysterectomy, with potential for functional imaging-based rehabilitation assessment and future research opportunities.
