Abstract

Cardiotocography records fetal heart rates and their temporal relationship to uterine contractions. To identify high risk fetuses, obstetricians inspect cardiotocograms (CTGs) by eye. Therefore, CTG traces are often interpreted differently among obstetricians, resulting in inappropriate interventions. However, few studies have focused on quantitative and nonbiased algorithms for CTG evaluation. In this study, we propose a newly constructed deep neural network model (CTG-net) to detect compromised fetal status. CTG-net consists of three convolutional layers that extract temporal patterns and interrelationships between fetal heart rate and uterine contraction signals. We aimed to classify the abnormal group (umbilical artery pH < 7.20 or Apgar score at 1 min < 7) and the normal group from CTG data. We evaluated the performance of the CTG-net with the F1 score and compared it with conventional algorithms, namely, support vector machine and k-means clustering, and another deep neural network model, long short-term memory. CTG-net showed the area under the receiver operating characteristic curve of 0.73 ± 0.04, which was significantly higher than that of long short-term memory. CTG-net, a quantitative and automated diagnostic aid system, enables early intervention for putatively abnormal fetuses, resulting in a reduction in the number of cases of hypoxic injury.

Highlights

  • Cardiotocography records fetal heart rates and their temporal relationship to uterine contractions

  • We demonstrated that deep neural network (DNN) models could predict infant outcomes from the last 30 min of CTG just before delivery

  • We clarified that the CTG-net model achieved significantly higher performance than the long short-term memory (LSTM) model

Read more

Summary

Introduction

Cardiotocography records fetal heart rates and their temporal relationship to uterine contractions. We propose a newly constructed deep neural network model (CTG-net) to detect compromised fetal status. CTG-net consists of three convolutional layers that extract temporal patterns and interrelationships between fetal heart rate and uterine contraction signals. The temporal relationship between FHR and UC has been widely considered to be a key factor for the interpretation of CTGs and categorized as early / variable / late / prolonged deceleration 2. One of the reasons is that previous studies have mainly focused on local wave patterns (extracted features thought to represent abnormal fetal status) and have not focused on postnatal o­ utcomes[10] Such humanannotated abnormal features of CTG are not necessarily consistent with adverse postnatal effects. There is little information available for adopting DNN in CTG interpretation, especially predicting infant outcomes, compared to those in other fields

Objectives
Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call