Robust Characterization of the Uterine Myoelectrical Activity in Different Obstetric Scenarios

Abstract

<jats:p>Electrohysterography (EHG) has been shown to provide relevant information on uterine activity and could be used for predicting preterm labor and identifying other maternal fetal risks. The extraction of high-quality robust features is a key factor in achieving satisfactory prediction systems from EHG. Temporal, spectral, and non-linear EHG parameters have been computed to characterize EHG signals, sometimes obtaining controversial results, especially for non-linear parameters. The goal of this work was to assess the performance of EHG parameters in identifying those robust enough for uterine electrophysiological characterization. EHG signals were picked up in different obstetric scenarios: antepartum, including women who delivered on term, labor, and post-partum. The results revealed that the 10th and 90th percentiles, for parameters with falling and rising trends as labor approaches, respectively, differentiate between these obstetric scenarios better than median analysis window values. Root-mean-square amplitude, spectral decile 3, and spectral moment ratio showed consistent tendencies for the different obstetric scenarios as well as non-linear parameters: Lempel–Ziv, sample entropy, spectral entropy, and SD1/SD2 when computed in the fast wave high bandwidth. These findings would make it possible to extract high quality and robust EHG features to improve computer-aided assessment tools for pregnancy, labor, and postpartum progress and identify maternal fetal risks.</jats:p>

Optimization of Imminent Labor Prediction Systems in Women with Threatened Preterm Labor Based on Electrohysterography

“…Spectral parameters are used to assess the shift of EHG spectral content to higher frequencies as labor approaches [ 12 , 17 , 32 ]. We computed the dominant frequencies in the bandwidth 0.2–1 Hz (DF1) and in 0.34–1 Hz (DF2), considering that the Fast Wave Low of the EHG ranges from 0.2 to 0.34 Hz and Fast Wave High, which includes EHG components above 0.34 Hz, limiting the high frequency to 1 Hz to diminish ECG and respiration interference [ 13 , 33 ]; the high-to-low frequency energy ratio (H/L ratio, as the ratio of energy in 0.34–1 Hz to energy in 0.2–0.34 Hz); deciles of the power spectrum and the spectral moment ratio (SMR) [ 20 , 34 ].…”

“…A set of representative complexity and regularity parameters to distinguish between preterm and term labor were also calculated to characterize the EHG signal, as in the related literature [ 20 ]: Lempel-Ziv (Binary and multistate n = 6), which assesses signal complexity from the number of different patterns in the signals [ 35 ]; sample, fuzzy and spectral entropy, which measure regularity based on the self-similarity in temporal and spectral domains [ 36 , 37 , 38 , 39 ]; and time reversibility [ 40 ], which estimates the similarity of a time series when time goes forward or back. A Poincare plot [ 41 ] of consecutive EHG signal samples was obtained and parameters SD1, SD2 were also computed to assess corresponding short- and long-term EHG changes, and SD1/SD2 to measure EHG randomness [ 42 ].…”

“…The EHG has been proposed as a promising technique for preterm labor prediction due to the fact that uterine activity increases its intensity and synchronization near labor [ 13 , 14 , 15 ]. The increased number of myometrial cells recruited in uterine contractions when labor is near results in a greater EHG amplitude, while the intensified cell excitability shifts the EHG spectral content to higher frequencies [ 16 , 17 ], and the labor onset entails changes in myometrial cell connectivity that modifies the regularity of the measured EHG signal: EHG predictability increases while signal complexity is reduced [ 18 , 19 , 20 ].…”

“…In the present study we focus on three different computationally efficient algorithms, random forest (RF), extreme learning machine (ELM) and K-nearest neighbors (KNN), which have been used in biological classification problems to distinguish between gestational or labor contractions or term/preterm deliveries from EHG recordings from routine checkups [ 28 ]. A recent study revealed that the 90th and 10th percentiles of the EHG parameters computed in 120 s whole analysis windows outperformed median values in discerning between different obstetric scenarios, as term vs. preterm labor, in recordings from routine checkups [ 20 ]. The better ability of the 10th–90th percentiles of EHG parameters to discriminate were associated with being more representative of the contractile segments in the recording session [ 20 ], choosing the 10th or 90th percentile of the EHG parameters according to their expected trend as delivery approaches [ 20 ].…”

“…A recent study revealed that the 90th and 10th percentiles of the EHG parameters computed in 120 s whole analysis windows outperformed median values in discerning between different obstetric scenarios, as term vs. preterm labor, in recordings from routine checkups [ 20 ]. The better ability of the 10th–90th percentiles of EHG parameters to discriminate were associated with being more representative of the contractile segments in the recording session [ 20 ], choosing the 10th or 90th percentile of the EHG parameters according to their expected trend as delivery approaches [ 20 ]. In addition, the optimization criterion in the classifier design was not clearly indicated in the literature regarding term/preterm birth or labor/nonlabor prediction, regardless of the classification algorithm employed [ 24 , 29 , 30 , 31 ].…”