According to the statement issued by University of California“Labour inductions are becoming increasingly common, but this does not mean that they end in vaginal birth, and in this sense the investigation was established.”Predicting vaginal delivery after induction of labor using machine learning“, written by Yolanda Ferreira, PhD student in Health Sciences. The doctoral thesis was supervised by Ana Luisa Arrea, Professor at FMUC, and co-supervised by João Nuno Correa, Professor at FCTUC.
“From the beginning, all inductions have a 30-35% chance of ending in a caesarean section, so we already know that 70% of women will deliver vaginally. However, if we could identify 30% of these cases that they would indeed end in caesarean section, we could provide adequate and proactive advice on the need to induce labour, which is a strenuous process for the mother and fetus which in turn “indeed, can add to the emotional and economic burden associated with this.” procedure,” explains Yolanda Ferreira.
So, “since this is a repetitive procedure, and produces a lot of data, we thought that maybe we could use a technology that would analyze it to help doctors understand if it is worth it or when it is worth investing in induction to achieve vaginal birth,” reveals the PhD student. The author highlights that obstetricians are currently investing in inducing labor in all women, knowing from the beginning that, due to certain characteristics, it may or may not occur vaginally.
Thus, according to João Nuno Correa, “the idea is to come up with something that, by combining data (tables and images), constitutes a support unit that provides personalized information for each pregnant woman, about the high probability of vaginal birth.” After induction.” “If this percentage is high, induction will be performed more confidently. If this is not the case, i.e. there is a very high possibility of a caesarean section, the pregnant woman may be advised otherwise.
“The innovation in this research is to predict the type of birth also using ultrasound image data. “The doctor relies on the person’s clinical history and characteristics after that pregnancy, and we want to see if the system, when analyzing that set of data and clinical images, looks at it in a way that Whether or not it later helps in arriving at a conclusion.” expose the investigators.
During the studies, researchers have already analyzed data from 2,600 women, followed at the Centro Hospitalar e Universitário de Coimbra (CHUC), which indicates promising results. The collected ultrasound will then be analyzed so that a tool can be created containing all the data that can be tested on real people.
“This collaboration between DEI and FMC is essential, because in the future it will support doctors’ prenatal decisions and thus allow for improved neonatal outcomes and women’s experiences during childbirth,” the researchers concluded.
CG
Related news
