Related Experiment Video
Updated: Oct 7, 2025
05:58
Application of an Amplitude-integrated EEG Monitor Cerebral Function Monitor to Neonates
Published on: September 6, 2017
39.5K
Network Theory Based EHG Signal Analysis and its Application in Preterm Prediction
IEEE Journal of Biomedical and Health Informatics
|January 5, 2022
Summary
This study uses electrohysterogram (EHG) signals to identify preterm birth risk. Network analysis of EHG data improves classification accuracy for term versus preterm pregnancies.
Area of Science:
- Biomedical Engineering
- Signal Processing
- Obstetrics
Background:
- Preterm birth is a major cause of infant mortality.
- Early detection of preterm birth risk is crucial for intervention.
- Uterine electrical activity is linked to contractions.
Purpose of the Study:
- To develop a precise method for classifying term and preterm pregnancies.
- To extract effective features from electrohysterogram (EHG) signals.
- To improve early identification of high-risk pregnancies.
Main Methods:
- Utilized Horizontal Visibility Graph (HVG) algorithm for network representation of EHG signals.
- Applied Short-Time Fourier Transform (STFT) for time-frequency domain analysis.
- Employed feature selection and Partition-Synthesis for imbalanced data to train Support Vector Machine (SVM) classifiers.
Main Results:
- Network-based features identified essential frequency components related to preterm birth.
- Achieved high classification performance with SVM: 0.89 sensitivity, 0.93 specificity, 0.91 accuracy, and 0.97 AUC.
- Demonstrated improved classification of term/preterm pregnancies.
Conclusions:
- Network analysis of EHG signals offers a promising approach for preterm birth prediction.
- The proposed method enhances the accuracy of identifying preterm birth risk.
- This technique can aid in timely medical interventions to prevent preterm birth.
