Machine Learning-Based Prediction Model of Preterm Birth Using Electronic Health Record.

Area of Science:

• Obstetrics and Gynecology
• Medical Informatics
• Machine Learning in Healthcare

Background:

• Preterm birth (PTB) is a leading cause of neonatal mortality.
• Early prediction of PTB is crucial for improving pregnancy outcomes.
• Machine learning offers potential for developing accurate PTB prediction models.

Purpose of the Study:

• To develop and evaluate a machine learning-based prediction model for PTB.
• To identify key factors influencing PTB risk.
• To assess the model's performance in early pregnancy prediction.

Main Methods:

• A retrospective study of 9550 pregnant women (2008-2018) from a Chinese hospital.
• Six machine learning algorithms (Naive Bayesian, SVM, Random Forest, ANN, K-means, Logistic Regression) were employed.
• Model performance was assessed using ROC curves, accuracy, sensitivity, and specificity.

Main Results:

• The Random Forest (RF) model demonstrated the highest accuracy (0.816) and AUC (0.885) at 27 weeks of gestation.
• RF model's accuracy and AUC showed a positive association with gestational age.
• Key PTB influencing factors included maternal age, magnesium, fundal height, and lipid profiles.

Conclusions:

• The RF-based prediction model shows significant potential for early PTB prediction.
• Interventions targeting identified PTB risk factors in early pregnancy may reduce incidence.
• Machine learning models can enhance obstetric care by enabling timely interventions.