First Trimester Prediction of Preterm Birth in Patient Plasma with Machine-Learning-Guided Raman Spectroscopy and Metabolomics

Preterm birth (PTB) is the leading cause of infant deathsglobally.Current clinical measures often fail to identify women who may deliverpreterm. Therefore, accurate screening tools are imperative for earlyprediction of PTB. Here, we show that Raman spectroscopy is a promisingtool for studying biological interfaces, and we examine differencesin the maternal metabolome of the first trimester plasma of PTB patientsand those that delivered at term (healthy). We identified fifteenstatistically significant metabolites that are predictive of the onsetof PTB. Mass spectrometry metabolomics validates the Raman findingsidentifying key metabolic pathways that are enriched in PTB. We alsoshow that patient clinical information alone and protein quantificationof standard inflammatory cytokines both fail to identify PTB patients.We show for the first time that synergistic integrationof Raman and clinical data guided with machine learning results inan unprecedented 85.1% accuracy of risk stratification of PTB in thefirst trimester that is currently not possible clinically. Correlationsbetween metabolites and clinical features highlight the body massindex and maternal age as contributors of metabolic rewiring. Ourfindings show that Raman spectral screening may complement currentprenatal care for early prediction of PTB, and our approach can betranslated to other patient-specific biological interfaces.

Automated summary

Preterm birth is a major cause of infant mortality globally. Current clinical measures often fail to identify women who may deliver preterm, making accurate screening tools essential. This study uses Raman spectroscopy to predict preterm birth by examining differences in the maternal metabolome. Fifteen statistically significant metabolites were identified as predictors of preterm birth. The integration of Raman and clinical data with machine learning achieved an unprecedented 85.1% accuracy in risk stratification of preterm birth in the first trimester.