Precision Unveiled in Unborn: A Cutting-Edge Hybrid Machine Learning Approach for Fetal Health State Classification.

Area of Science:

• Medical informatics
• Computational biology
• Maternal-fetal medicine

Background:

• Fetal health classification is crucial for improving maternal and infant outcomes.
• Early detection of fetal cardiac abnormalities aids timely prenatal management.
• Machine learning (ML) algorithms, including deep learning, are advancing fetal electrocardiogram (ECG) signal analysis.

Purpose of the Study:

• To develop and evaluate a novel hybrid machine learning model for enhanced fetal health classification.
• To improve the accuracy and reliability of fetal monitoring and disease prediction during pregnancy.
• To leverage ML for automated decision-making in fetal health emergencies and telemonitoring.

Main Methods:

• A hybrid approach combining Random Forest (RF) and AdaBoost algorithms was developed.
• The model integrates RF's capability with high-dimensional data and AdaBoost's focus on classification accuracy.
• Existing models and challenges in fetal health data handling were reviewed to inform the hybrid model design.

Main Results:

• The hybrid model achieved 95.98% classification accuracy on a benchmark CTG dataset.
• High precision (92.88%), recall (92.78%), and F1 score (92.70%) were obtained.
• The results indicate the model's potential for real-world applications in early fetal anomaly detection.

Conclusions:

• A novel hybrid RF-AdaBoost model offers accurate and reliable fetal health classification.
• The combined approach demonstrates superior performance and robustness over standalone models.
• This model represents a significant contribution to the health sector for timely fetal-maternal health evaluation and prediction.