Linear and non-linear features for intrapartum cardiotocography evaluation

Since 1960's obstetricians have been using cardiotocography (CTG) to detect possible ongoing hypoxia of the fetus. CTG consists of fetal heart rate (fHR) and uterine contraction (TOCO) monitoring. The evaluation of the fHR in clinical settings is ruled by FIGO guidelines, which are based on evaluation of macroscopic morphological features derived from the fHR, such as baseline variability. Although upgrades were proposed to the guidelines - none of them is taking into account results achieved by the adult heart rate variability research. In this work, almost complete set of features previously used for fHR description is investigated and the features are assessed based on their statistical significance in the task of distinguishing the records into three FIGO classes. Inter-correlation of the features is also discussed. We assess the features on a large data set and use expert signal evaluation instead of pH values with the aim to give an overall view of the potential usefulness of the features in the clinical settings. We conclude the paper by presenting the best uncorrected feature subset according to the meta-analysis of three different ranking methods.