Scholarlink

Background <p>Ectopic Pregnancy (EP) is a type of pregnancy where the developing blastocyst implants in a location other than the endometrial cavity. The aim of study was developing predictive models that could improve the accuracy of identifying individuals at risk and uncover relationships between known risk factors and ectopic pregnancy occurrence.</p> Methods <p>Five-fold cross-validation were employed to prevent overfitting and the Grid Search method was utilized to determine the optimal hyper-parameters for the models. The performance of the models was evaluated using metrics such as accuracy, Area Under the Curve (AUC), and Negative Predictive Value (NPV). The SHapley Additive exPlanations (SHAP) method was used to interpret the model’s decision-making process and identify the most significant features.</p> Results <p>RF demonstrated the best performance (87.13% accuracy, 90.65% AUC). Key predictors identified via logistic regression (LR) and SHAP analysis included mid-cycle pain, genital surgery history and dysmenorrhea.</p> Conclusion <p>This investigation illustrates the potential for machine learning models to improve clinical decision-making and optimize patient outcomes in cases of ectopic pregnancy. Further validation in diverse populations is required due to the constraints of single-center data and the absence of critical risk factors, such as the use of PIDs and IUDs.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of ectopic pregnancy using interpretable machine learning algorithms

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月21日

查看全文
  • Arkan Aghayari,
  • Amir Sorayaie Azar,
  • Mortaza Taheri-Anganeh,
  • Sonia Sadeghpour,
  • Yousef Mohammadpour,
  • Jamshid Bagherzadeh Mohasefi,
  • Hojat Ghasemnejad-Berenji

摘要

Background

Ectopic Pregnancy (EP) is a type of pregnancy where the developing blastocyst implants in a location other than the endometrial cavity. The aim of study was developing predictive models that could improve the accuracy of identifying individuals at risk and uncover relationships between known risk factors and ectopic pregnancy occurrence.

Methods

Five-fold cross-validation were employed to prevent overfitting and the Grid Search method was utilized to determine the optimal hyper-parameters for the models. The performance of the models was evaluated using metrics such as accuracy, Area Under the Curve (AUC), and Negative Predictive Value (NPV). The SHapley Additive exPlanations (SHAP) method was used to interpret the model’s decision-making process and identify the most significant features.

Results

RF demonstrated the best performance (87.13% accuracy, 90.65% AUC). Key predictors identified via logistic regression (LR) and SHAP analysis included mid-cycle pain, genital surgery history and dysmenorrhea.

Conclusion

This investigation illustrates the potential for machine learning models to improve clinical decision-making and optimize patient outcomes in cases of ectopic pregnancy. Further validation in diverse populations is required due to the constraints of single-center data and the absence of critical risk factors, such as the use of PIDs and IUDs.