Pregnancy/Labor Discrimination and Monitoring: An Investigation Using Nonlinear Methods

Abstract

In the last ten years the ratio of preterm labor mortality increases. Many studies were done about the cause of gestation as well as preterm labor. Nowadays, studies are dealing with the signal detection of uterus contraction which can be used by analysis methods in order to determine its features and its role in labor. In this study, many Non-linear methods are used to extract features from the signal in order to differentiate between pregnancy and labor signals and to monitor pregnancy of women for each week before labor (WBL): Lempel-Ziv complexity (lzc), Fractal Dimension (FD), Hjorth parameter. All data recorded between Lebanon and France from 12 WBL until 1 WBL and labor using a 4×4 matrix of electrodes. Methods were tested first on synthetic signals to test their sensitivity to nonlinearity change then they were applied on real signals. Results show the implementation of the nonlinear analysis methods of signal processing to differentiate between the contraction signals, which appears as the variation and sensitivity of pregnancy and labor signals with respect to methods that can help to detect normal and preterm labor.