Preterm Labor:
Innovative research by Singh et al., studied the combination of AI and
amniotic fluid (AF) proteomics and metabolomics, in conjunction or
independently with imaging, demographic and clinical factors, to predict
perinatal outcome in asymptomatic women with short cervix length
[14]. The type of AI they used is called deep learning (DL). This
subtype of AI can operate with a larger amount of data because it has a
greater number of neural networks, which makes it ideal for biological
system studies involving multiomics. Currently, the short cervical
length is the strongest risk factor for prematurity, however, many women
with this condition carry their pregnancy to term. Many centers now
incorporate amniocentesis in these women to evaluate additional factors
that might put them at risks, such as inflammation and infectious
processes. The AF of the subjects was additionally studied for omics,
such as metabolomics, to shed light on potential new biomarkers that
might be involved in preterm birth. This can improve the accuracy and
predictive value of women at risk of poor outcomes, and it can help
physicians stratify those patients at risk of preterm birth better than
the current risk factors such as short cervical length and prior preterm
birth delivery [15]. In this way, physicians can use this tool to
guide their management, such as observation alone, or suggesting
cervical cerclage and or antenatal steroids if deemed necessary. A
shortcoming of the study was the small size of the group. The study
concluded that DL was a superior tool when it came to the prediction of
perinatal outcome in asymptomatic women with short cervix length and
that further studies are needed to analyze AF omics and its relationship
to premature shortening of the cervix to help guide management in these
patients.
A study done by Idowu et al, emphasizes the importance of using AI
technology to decrease expenses generated by inaccurate detection of
preterm labor leading to unnecessary hospitalizations and procedures,
and at the meantime, expedite treatment in those who are in true labor
to prevent hazardous consequences for the baby and the mother [16].
In this study, they used electro-hysterography signals and used three
distinct machine learning algorithms to classify these signals to help
them identify true labor and accurately diagnose preterm labor. They
concluded that the Random Forest algorithm performed the most
efficiently of the three machines tested, as it was able to handle a
larger amount of data, is relatively accurate, and has a robust learning
capacity which resulted in an accuracy of 97 percent in predicting
preterm labor.