Scholarlink

In Mexico, the maternal mortality rate stands at approximately 59 deaths per 100,000 live births due to pregnancy-related causes. Early identification of gestational risks is crucial for preventing complications during and after pregnancy that can impact both mother and child. These risks may arise from various factors, including lifestyle, age, obesity, and preexisting medical conditions such as diabetes. However, early detection is often challenging due to the complexity and variability of symptoms across individuals. This highlights the need for automation tools, such as machine learning models, to efficiently process data and identify patterns that effectively assess gestational risks. Therefore, in this work, we propose the application of three machine learning algorithms: Artificial Neural Network (ANN), Decision Tree (DT), and Random Forest (RF), to predict the maternal risk during pregnancy. The data used in this study includes 19 clinical features of Mexican women, classified in three different states: healthy, risky, and possible death. The results showed that all algorithms achieved over 90% accuracy, with the Random Forest model obtaining the best performance, with an average F1-score of 95.19%. This underscores the potential of machine learning techniques to significantly improve early diagnosis and intervention strategies, thereby increasing the likelihood of safer pregnancy outcomes.

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

Maternal Risk Prediction During Pregnancy Through Machine Learning Using Mexican Women’s Data

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Roberto Hernández-Chávez,
  • Yair L. Grijalva-González,
  • Bernardo O. Enriquez-Guillen,
  • Javier Camarillo-Cisneros,
  • Natalia G. Sámano-Lira,
  • Abimael Guzman-Pando

摘要

In Mexico, the maternal mortality rate stands at approximately 59 deaths per 100,000 live births due to pregnancy-related causes. Early identification of gestational risks is crucial for preventing complications during and after pregnancy that can impact both mother and child. These risks may arise from various factors, including lifestyle, age, obesity, and preexisting medical conditions such as diabetes. However, early detection is often challenging due to the complexity and variability of symptoms across individuals. This highlights the need for automation tools, such as machine learning models, to efficiently process data and identify patterns that effectively assess gestational risks. Therefore, in this work, we propose the application of three machine learning algorithms: Artificial Neural Network (ANN), Decision Tree (DT), and Random Forest (RF), to predict the maternal risk during pregnancy. The data used in this study includes 19 clinical features of Mexican women, classified in three different states: healthy, risky, and possible death. The results showed that all algorithms achieved over 90% accuracy, with the Random Forest model obtaining the best performance, with an average F1-score of 95.19%. This underscores the potential of machine learning techniques to significantly improve early diagnosis and intervention strategies, thereby increasing the likelihood of safer pregnancy outcomes.