Construction of a predictive model for the risk of cesarean scar diverticula formation after cesarean delivery based on multifactorial Logistic regression
10.19405/j.cnki.issn1000-1492.2025.07.019
- VernacularTitle:基于多因素 Logistic 回归构建剖宫产术后子宫 疤痕憩室形成风险预测模型
- Author:
Mengyuan Zhang
1
;
Ye He
1
;
Yuanyuan Wu
2
;
Jing Wang
1
Author Information
1. Dept of Obstetrics and Gynecology，The First Affiliated Hospital of Anhui Medical University，Hefei 230022
2. Dept of Ultrasound，The First Affiliated Hospital of Anhui Medical University，Hefei 230022
- Publication Type:Journal Article
- Keywords:
cesarean scar diverticula;
machine learning;
LASSO cross-validation;
risk prediction;
cesarean section;
nomogram
- From:
Acta Universitatis Medicinalis Anhui
2025;60(7):1297-1304
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To screen the risk factors of cesarean scar diverticula ( CSD) after cesarean section and to construct a risk prediction model．
Methods:491 cases of mothers who underwent cesarean section were recruited as the study subjects，and the data from the database of negative ultrasound of mothers who returned to the hospital 12 months after operation were collected，and the dataset was randomly divided into the training set and the test group according to 7 ∶ 3 ; the variables were screened to obtain the risk factors of CSD and the risk prediction model was constructed by the use of least absolute shrinkage and selection operator (LASSO) ; the variables were screened using the LASSO to obtain the characteristic variables，and the characteristic variables were analyzed by multifacto- rial logistic regression analysis，and the nomogram prediction model was constructed by using the R software．
Results:A total of 491 cases of sample data were included，including 344 cases in the training set and 147 cases in the test set ; feature variables were screened by LASSO，and ten－fold cross－validation was used．Five variables were finally screened : number of cesarean deliveries，number of years between two cesarean deliveries，24-hour hemorrhage，operation time and uterine position(P＜0. 05) ．The accuracy of the decision analysis curves for inter- nal evaluation and internal validation of the CSD risk prediction model constructed using it was high ; the AUC (95% CI) of the diagnostic model in the training set and the test set were 0. 75 (0. 71－0. 80) and 0. 79 (0. 71 - 0. 87) ，respectively．
Conclusion :The risk prediction model established using the LASSO cross-validation algo- rithm has good predictive value for the occurrence of postpartum scar diverticula，which deserves clinical attention．
- Full text:2026041422590403147基于多因素Logistic回归构建剖宫产术后子宫疤痕憩室形成风险预测模型_张梦媛.pdf
