王瑜(共同一作),中国人民大学统计学院副教授,北京生物医学统计与数据管理研究会副秘书长、国家食品药品监督管理总局医疗器械技术审评专家咨询委员会委员、中华预防医学会健康保险专业委员会委员、北京医学会健康管理学分会质量控制与改进学组委员、中国健康促进与教育协会健康管理分会委员。主要从事医疗保障、健康调查、老龄化、指数分析、医疗数据分析等方面的研究。研究论文发表在Journal of Affective Disorders、Public Health、BMC Public Health、 Scientific Reports、PLoS One、BMC Health Services Research等杂志上。主持了国家社会科学基金项目、中国人民大学研究品牌计划基础研究项目等。
Sequential three-dimensional ultrasound diagnosis of a T-shaped uterus
4
报告摘要
Objectives:
To develop a pragmatic and precise diagnostic approach for a T-shaped uterus.
Study design:
This retrospective case-control study was conducted at a national referral center for congenital uterine malformations in Beijing, China, with cases enrolled between July 2021 and June 2024 and controls recruited between August and November 2025. Fifty patients with a T-shaped uterus were identified as the case group and diagnosed by a panel of 12 experts through blinded independent assessment of three-dimensional transvaginal ultrasonography (3D-TVS) and hysteroscopic findings (consensus, ≥7/12). A control group of 50 fertile women with functionally normal uteri was prospectively recruited. All ultrasound examinations were performed during the secretory phase. Thirteen quantitative sonographic parameters were analyzed. Least Absolute Shrinkage and Selection Operator (LASSO) regression was used to identify candidate predictors. Based on LASSO coefficients and clinical relevance, three key parameters-cavity width at the level of lateral wall indentation, average lateral internal indentation depth, and upper cavity depth-were selected for the diagnostic model. Single-parameter and sequential (two- and three-step) strategies were evaluated using receiver operating characteristic (ROC) analysis. Sequential strategies used “grey zones” defined by 100% sensitivity or specificity thresholds in the initial steps, with final arbitration based on Youden’s index. Diagnostic performance was assessed using sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV), and area under the curve (AUC).
Results:
Reference values for 13 secretory-phase ultrasound parameters in functionally normal uteri were established, including cavity width at the level of lateral wall indentation (95% confidence interval [CI]: 15.0–16.9 mm), average lateral internal indentation depth (95% CI: 3.9–4.6 mm), and upper cavity depth (95% CI: 13.3–15.6 mm). For single-parameter diagnosis, parameter A (cavity width ≤14 mm) achieved 100% sensitivity and 72% specificity; parameter B (indentation depth ≥5 mm) showed 100% sensitivity and 76% specificity; and parameter C (upper cavity depth ≤13 mm) yielded 98% sensitivity and 60% specificity. The two-step sequential strategy (A + B) improved specificity to 96% with a sensitivity of 90%. The proposed three-step sequential strategy (A + B + C) employed a grey-zone approach: step 1 (width 10–14 mm) triaged cases; step 2 (depth 5–7 mm) resolved ambiguity; and step 3 (upper depth ≤12 mm) confirmed the diagnosis. This combined strategy showed the highest diagnostic performance, with a sensitivity of 98%, specificity of 96%, accuracy of 97%, PPV of 96%, and NPV of 98%.
Conclusions:
The proposed 3D-TVS-based three-step sequential strategy provides an objective, accurate, and reproducible diagnostic approach for T-shaped uteri, effectively reducing diagnostic ambiguity.