Related Experiment Video
Updated: Oct 4, 2025
A Syngeneic Murine Model of Endometriosis using Naturally Cycling Mice
Published on: November 24, 2020
Evaluation of menopausal endometrial lesions via mathematical modeling clinical indicators and ultrasonographic
Background:
The diagnosis of benign and malignant menopausal endometrial lesions (MEL) is often misled by complicated clinical indicators and ultrasonographic parameters in actual clinical applications.
Objective:
To investigate the performance of clinical indicators and ultrasonographic parameters in the diagnosis of MEL.
Methods:
A cohort of 156 enrolled menopausal patients with MEL was divided into benign group (128 cases) and malignant group (28 cases). Two clinical indicators of patient age (PA), abnormal vaginal bleeding (AVB) and three transvaginal ultrasonography (TVS) parameters of endometrial thickness (ET), endometrial uneven echo (EUE) and endometrial blood flow signal (EBFS) were measured for the mathematical modelling. The performance of combined indicators and individual indicators were firstly compared, and then the optimized combined indicators was compared with corresponding individual indicators, respectively.
Results:
Our experiments verified that the mathematical modelling presented robust capabilities in the diagnosis of MEL with the sensitivity, specificity and AUC of 78.6%, 75.8% and 0.83 for combined indicators, and 75.0%, 81.3% and 0.85 for optimized combined indicators, respectively. The cut off thresholds of PA was 57.5 years, ET was 11.5 mm. Furthermore, the AVB presented the most important risk factor among the optimized indicators of PA, ET and AVB (P< 0.05).
Conclusions:
The combined indicators presented better performance in differentiating benign and malignant MEL and the AVB demonstrated the most capability for clinical applications.
