Bosque Aleatorio Basado en Rasgos Tiempo-Frecuencia de la VFC Fetal para la Identificación de Actividad Uterina en Fetos a Término y Pretérmino

Abstract

The aim of this work was to discriminate periods of contraction and periods of rest during labor by implementing a random forest classifier based on time-frequency features of the fetal heart rate variability (fHRV) from term and preterm fetuses. fHRV signals from term (PT = 38) and preterm (PP = 25) fetuses were analyzed during active labor. The signals were segmented with respect to the presence or absence of uterine activity. Subsequently, time-frequency indices of fHRV such as flow (0 °, 45 ° and 90 °) and energy (very low, low and high frequency) were calculated. The classifier obtained an area under the ROC curve of 0.87 and 0.88, for PT and PP respectively. Our results suggest that for both groups (PT and PP), there is a different cardiac dynamic between periods of contraction and periods of rest. Therefore, it is important to distinguish between the two periods to adequately characterize the fetal response to labor.