Parameter-based analysis of antepartum and labour electrohysterogram signals

Monitoring of the evolution of the pregnancy from early stages presents high interest due to the possibility of identifying high risk situations that can be further managed by health care professionals. One of the more recent and promising methods of monitoring fetal and maternal physiological wellbeing is the electrohysterogram (EHG) which is a non-invasive technique of acquiring uterine electrical activity. The purpose of this paper is the analysis of EHG derived parameters for both pregnancy (antepartum) and labour recordings in order to identify significant statistical differences. The analysis focuses on time domain features represented by four parameters and frequency domain features represented by three parameters and identifies the features that can be used as means of classification based on highest difference between pregnancy and labour recordings. The recordings analysed are obtained using a 4×4 grid of sensors placed on the patient's abdomen resulting in 16 channels signal.