Dispersion Entropy: A Measure of Electrohysterographic Complexity for Preterm Labor Discrimination

Nieto-del-Amor, Félix; Ye-Lin, Yiyao; Garcia-Casado, Javier; Díaz-Martinez, Alba; Monfort-Ortíz, Rogelio; Prats-Boluda, Gema

doi:10.5220/0010316602600267

Cited by 4 publications

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…After Vandewiele et al corrected this problem, these models were no longer able to predict preterm births accurately [21,22]. Additional works with sound methodology suggest that some features derived from EHG measurements can be used to distinguish between recordings of mothers who eventually delivered at term from those who delivered preterm [23,24,25,26]. However, none of these works could predict preterm births with clinically useful accuracy.…”

Section: Introductionmentioning

confidence: 99%

Predicting preterm births from electrohysterogram recordings via deep learning

Goldsztejn

Nehorai

2022

Preprint

View full text Add to dashboard Cite

About one in ten babies is born preterm, i.e., before completing 37 weeks of gestation, which can result in permanent neurologic deficit and is a leading cause of child mortality. Although imminent preterm labor can be detected, predicting preterm births more than one week in advance remains elusive. Here, we develop a deep learning method to predict preterm births directly from electrohysterogram (EHG) recordings of pregnant mothers without symptoms of preterm labor. We developed a prediction model, which includes a recurrent neural network, to predict preterm births using short-time Fourier transforms of EHG recordings and clinical information from two public datasets. We predicted preterm births with an area under the receiver-operating characteristic curve (AUC) of 0.80 (95% confidence interval: 0.79-0.82). Moreover, we found that the spectral patterns of the measurements were more predictive than the temporal patterns, suggesting that preterm births can be predicted from short EHG recordings in an automated process. We show that, even without symptoms of preterm labor, preterm births can be predicted for pregnant mothers around their 31st week of gestation, prompting beneficial treatments to reduce the incidence of preterm births and improve their outcomes.

show abstract