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

Novel machine learning applications in peripartum care: a scoping review

OBJECTIVE

Machine learning (ML), a subtype of artificial intelligence (AI), presents predictive modeling and dynamic diagnostic tools to facilitate early interventions and improve decision-making. Considering the global challenges of maternal, fetal, and neonatal morbidity and mortality, ML holds the potential to enable significant improvements in maternal and neonatal health outcomes. We aimed to conduct a comprehensive review of ML applications in peripartum care, summarizing the potential of these tools to enhance clinical decision-making and identifying emerging trends and research gaps.

DATA SOURCES

We conducted a scoping review on MEDLINE, Cochrane Library, and EMBASE databases from inception to April 2024. We gathered additional relevant studies through snowball sampling. We meticulously screened titles and abstracts and chose full-text articles for further analysis.

STUDY ELIGIBILITY CRITERIA

We included primary research articles and abstracts focusing on pregnant individuals, employing ML methods for peripartum care.

STUDY APPRAISAL AND SYNTHESIS METHODS

No formal quality assessment was performed. Data were extracted using a custom template to capture study characteristics and ML models. Findings were synthesized using summary tables and figures to highlight key trends and results.

RESULTS

Among 406 studies, 78% were published within the last five years. Most studies originated from high-income or well-resourced countries, with 27% from North America (including 24% from the United States) and 34% from Asia, predominantly China (18%). Studies from low- and middle-income regions were notably scarce, reflecting significant regional disparities. Predictive modeling tasks were the most prevalent (59%), followed by classification tasks (29%). Supervised learning dominated (90%), with algorithms such as Support Vector Machines, Random Forests, and Logistic Regression most commonly used. Key topics included fetal distress and acidemia (32%), preterm birth (22%), mode of delivery (13%), and birth weight (13%). Notably, Explainable AI methods were utilized in only 19% of studies, and external validation was performed in just 5%. Despite these advancements, only 1% of models resulted in accessible clinical tools, and none were fully integrated into healthcare systems.

CONCLUSIONS

ML holds significant potential to enhance peripartum care by improving diagnostic accuracy and predictive capabilities. However, realizing this potential requires responsible AI practices, including robust validation with external datasets, prospective investigations across diverse populations, and the development of digital and data infrastructure for seamless integration into electronic health records. Additionally, transparent AI that provides insights into risk stratification logic is essential for clinician trust in ML tools. Future research should address understudied areas, prioritize neglected low-income settings, and explore advanced ML approaches to improve maternal and neonatal outcomes.

The single-cell immune profile throughout gestation and its potential value for identifying women at risk for spontaneous preterm birth

Precisely timed immune adaptations, observed in the maternal circulation, underpin the notion of an immune clock of human pregnancy that supports its successful progression and completion at delivery. This immune clock is divided into three immunological phases, with the first phase starting at the time of conception and implantation, shifting into the second phase that supports homeostasis and tolerance throughout pregnancy, and culminating in the last phase of labor and parturition. Disruptions of this immune clock are reported in pregnancy complications such as spontaneous preterm birth. However, our understanding of the immune clock preceding spontaneous preterm birth remains scattered. In this review, we describe the chronology of maternal immune cell adaptations during healthy pregnancies and highlight its disruption in spontaneous preterm birth. With a focus on single-cell cytometric, proteomic and transcriptomic approaches, we review recent studies of term and spontaneous preterm pregnancies and discuss the need for future prospective studies aimed at tracking pregnancies longitudinally on a multi-omic scale. Such studies will be critical in determining whether spontaneous preterm pregnancies progress at an accelerated pace or follow a preterm-intrinsic pattern when compared to those delivered at term.

Enzymatic Desialylation Enables Reliable Charge Variant Characterization of Highly Glycosylated and Sialylated Fc Fusion Proteins

Fusion proteins constitute a class of engineered therapeutics and have emerged as promising candidates for disease treatment. However, the structural complexity and heterogeneity of fusion proteins make their characterization extremely challenging, and thus, an innovative and comprehensive analytical toolbox is needed. Here, for the first time, we demonstrate a novel and robust workflow to evaluate charge variants for a highly glycosylated fusion protein with heavy sialylation using imaged capillary isoelectric focusing (icIEF). In the development of the icIEF method, key factors that were systematically investigated include the desialylation level, the stability of the desialylated molecule, incubation time and temperature of desialylation, protein concentrations, urea and l-arginine effects on the tertiary structure, and instrumental comparability. Multivariate and correlation analyses were subsequently applied to confirm the impacts of the parameters evaluated. Furthermore, a microfluidic chip-based icIEF system coupled with ultraviolet detection and mass spectrometry (icIEF-UV/MS) was utilized to identify critical post-translational modifications and ameliorate the understanding of charge variants. Our study demonstrates that this workflow enables a mechanistic understanding of charge variants for heavily sialylated therapeutics.

Monitoring kinetic processes of drugs and metabolites: Surface-enhanced Raman spectroscopy

Monitoring the kinetic changes of drugs and metabolites plays a crucial role in fundamental research, preclinical and clinical application. Raman spectroscopy (RS) is regarded as a fingerprinting technique that can reflect molecular structures but limited in applications due to poor sensitivity. Surface-enhanced Raman spectroscopy (SERS) significantly amplifies the detection sensitivity by plasmonic substrates, facilitating the identification and quantification of small molecules in biological samples, such as serum, urine, and living cells. This review will focus on advances in how SERS has been utilized to monitor the dynamic processes of small molecule drugs and metabolites in recent years. We first provide readers with a comprehensive overview of the mechanism and practical considerations of SERS, including enhancement theory, substrate design, sample pretreatment, molecule-substrate interactions and spectral analysis. Then we describe the latest advances in SERS for the detection and analysis of metabolites and drugs in cells, dynamic monitoring of drug in various biological matrices, and metabolic profiling for health assessment in biological fluids. We believe that high-performance SERS substrates, standardized technical regulations, and artificial intelligence spectral analysis will boost sensitive, accurate, reproducible, and universal molecular detection in the future. We hoped this review could inspire researchers working in related fields to better understand and utilize SERS for the analytical detection of drugs and metabolites.

Discrimination of Benign and Malignant Thyroid Nodules through Comparative Analyses of Human Saliva Samples via Metabolomics and Deep-Learning-Guided Label-free SERS

Thyroid nodules are a very common entity. The overall prevalence in the populace is estimated to be around 65-68%, among which a small portion (less than 5%) is malignant (cancerous). Therefore, it is important to discriminate benign thyroid nodules from malignant thyroid nodules. In this study, an equal number of participants with benign and malignant thyroid nodules (N = 10/group) were recruited. Saliva samples were collected from each participant, and SERS spectra were acquired, followed by validation using a metabolomics approach. An additional equal number of patients (N = 40/group) were recruited to construct diagnostic models. The performance of various machine learning (ML) algorithms was assessed using multiple evaluation metrics. Finally, the reliability of the optimal model was tested using blind test data (N = 10/group for benign and malignant thyroid nodules). The results showed a consistent trend between the SERS metabolic profile and the metabolites identified through MS analysis. The Multi-ResNet algorithm was optimal, achieving a 95% accuracy in sample discrimination. Additionally, blind test data sets yielded an overall accuracy of 83%. In summary, the deep-learning-guided SERS technique holds great potential in the accurate discrimination of benign and malignant thyroid nodules via human saliva samples, which facilitates the noninvasive diagnosis of malignant thyroid nodules in clinical settings.

Rapid diagnosis of bacterial vaginosis using machine-learning-assisted surface-enhanced Raman spectroscopy of human vaginal fluids

Bacterial vaginosis (BV) is an abnormal gynecological condition caused by the overgrowth of specific bacteria in the vagina. This study aims to develop a novel method for BV detection by integrating surface-enhanced Raman scattering (SERS) with machine learning (ML) algorithms. Vaginal fluid samples were classified as BV positive or BV negative using the BVBlue Test and clinical microscopy, followed by SERS spectral acquisition to construct the data set. Preliminary SERS spectral analysis revealed notable disparities in characteristic peak features. Multiple ML models were constructed and optimized, with the convolutional neural network (CNN) model achieving the highest prediction accuracy at 99%. Gradient-weighted class activation mapping (Grad-CAM) was used to highlight important regions in the images for prediction. Moreover, the CNN model was blindly tested on SERS spectra of vaginal fluid samples collected from 40 participants with unknown BV infection status, achieving a prediction accuracy of 90.75% compared with the results of the BVBlue Test combined with clinical microscopy. This novel technique is simple, cheap, and rapid in accurately diagnosing bacterial vaginosis, potentially complementing current diagnostic methods in clinical laboratories.

IMPORTANCE

The accurate and rapid diagnosis of bacterial vaginosis (BV) is crucial due to its high prevalence and association with serious health complications, including increased risk of sexually transmitted infections and adverse pregnancy outcomes. Although widely used, traditional diagnostic methods have significant limitations in subjectivity, complexity, and cost. The development of a novel diagnostic approach that integrates SERS with ML offers a promising solution. The CNN model's high prediction accuracy, cost-effectiveness, and extraordinary rapidity underscore its significant potential to enhance the diagnosis of BV in clinical settings. This method not only addresses the limitations of current diagnostic tools but also provides a more accessible and reliable option for healthcare providers, ultimately enhancing patient care and health outcomes.

Associations of air pollutants and related metabolites with preterm birth during pregnancy

OBJECTIVE

This study aimed to investigate the influence of exposure to ambient fine particulate matter (PM2.5) and its components during pregnancy on the prevalence of preterm birth (PTB). Additionally, we sought to identify the susceptible exposure window. Furthermore, we explored the potential mediating role of blood analysis and a comprehensive metabolic panel in the association between pollutant exposure and PTB incidence.

METHODS

This birth cohort study recruited 139 participants with PTB outcomes and 1713 controls from Fujian Maternal and Child Health Hospital between January 2021 and June 2023. Sociodemographic characteristics and clinical treatment data during participants' first pregnancies were collected. The exposure levels to pollutants during pregnancy were estimated via a combined geographic-statistical model utilising satellite remote sensing data. The distributional lag nonlinear modelling was employed to assess associations between pollutant exposure during pregnancy and the prevalence of PTB. Weighted quantile regression was used to identify key components associated with PM2.5 and PTB during pregnancy. Additionally, a mediating effect analysis was conducted to evaluate the role of blood analysis. The metabolic profile was used to screen for differentially abundant metabolites associated with PTB and explore their relative expression in relation to air pollutants and PTB incidence.

RESULTS

Following the adjustment for potential confounding variables, the mean weekly susceptibility windows for PM2.5 were identified as 7-10, 16-19, and 22-28 weeks; 8-10, and 15-19 weeks for inorganic sulfate; 6-10, and 15-28 weeks for nitrate; 6-12, and 15-28 weeks for ammonium (NH4+); and 7-9, 18-20, and 22-36 weeks for organic matter. During mixed exposure to PM2.5 components, the key component is NH4+. In the mixed exposure to PM2.5 components, NH4+ emerged as a key contributor. The results of the mediation analysis revealed that haemoglobin played a mediating role, accounting for 21.53 % of the association between exposure to environmental pollutants and the prevalence of PTB. It is noteworthy that, no mediating effects were observed for the other variables. Furthermore, non-targeted metabolomics identified 17 metabolites associated with PTB. Among these factors, hydrogen phosphate may impact metabolic pathways such as oxidative phosphorylation, influencing the risk of PTB. The interplay between environmental pollutants and metabolites, particularly through oxidative phosphorylation pathways, may contribute to PTB incidence.

CONCLUSIONS

The evidence indicates that exposure to PM2.5 and its components during pregnancy were a significant risk factor for PTB. Notably, specific weekly exposure windows were identified for pollutants during pregnancy. Among the PM2.5 components, NH4+ exhibited the most substantial weight in the association analysis between exposure to the mixture of components and PTB. Furthermore, our mediation analysis revealed that haemoglobin serves as a partial mediator in the relationship between exposure to pollutants during pregnancy and the prevalence of PTB. Additionally, maternal serum metabolic profiles differed between the preterm and control groups. Notably, a combined effect involving hydrogen phosphate and mixed exposure to PM2.5 fractions further contributed to the development of PTB. Oxidative phosphorylation pathways may play pivotal roles in this intricate association.

Diagnosis of pregnancy disorder in the first-trimester patient plasma with Raman spectroscopy and protein analysis

Gestational diabetes mellitus (GDM) is a pregnancy disorder associated with short- and long-term adverse outcomes in both mothers and infants. The current clinical test of blood glucose levels late in the second trimester is inadequate for early detection of GDM. Here we show the utility of Raman spectroscopy (RS) for rapid and highly sensitive maternal metabolome screening for GDM in the first trimester. Key metabolites, including phospholipids, carbohydrates, and major amino acids, were identified with RS and validated with mass spectrometry, enabling insights into associated metabolic pathway enrichment. Using classical machine learning (ML) approaches, we showed the performance of the RS metabolic model (cross-validation AUC 0.97) surpassed that achieved with patients' clinical data alone (cross-validation AUC 0.59) or prior studies with single biomarkers. Further, we analyzed novel proteins and identified fetuin-A as a promising candidate for early GDM prediction. A correlation analysis showed a moderate to strong correlation between multiple metabolites and proteins, suggesting a combined protein-metabolic analysis integrated with ML would enable a powerful screening platform for first trimester diagnosis. Our study underscores RS metabolic profiling as a cost-effective tool that can be integrated into the current clinical workflow for accurate risk stratification of GDM and to improve both maternal and neonatal outcomes.

Metabolic Response to Small Molecule Therapy in Colorectal Cancer Tracked with Raman Spectroscopy and Metabolomics

Despite numerous screening tools for colorectal cancer (CRC), 25 % of patients are diagnosed with advanced disease. Novel diagnostic technologies that are early, accurate, and rapid are imperative to assess the therapeutic efficacy of clinical drugs and identify new biomarkers of treatment response. Here Raman spectroscopy (RS) was used to track metabolic reprogramming in KRAS-mutant HCT116 and SW837 cells, and KRAS wild-type CC cells. RS combined with multivariate analysis methods distinguished nonresponsive, partially responsive, and responsive cells treated with cetuximab, a monoclonal antibody for EGFR inhibition, sotorasib, a clinically approved KRAS inhibitor, and various doses of trametinib, an inhibitor of the MAPK pathway. Cells treated with a combination of subtoxic doses of trametinib and BKM120, an inhibitor of the PI3K pathway, showed a synergistic response between the two pathways. Using a supervised machine learning regression model, we established a scoring methodology trained to a priori predict therapeutic response to new treatment combinations. RS metabolites were verified with mass spectrometry, and enrichment pathways were identified, including amino acid, purine, and nicotinate and nicotinamide metabolism that differentiated monotherapy from combination therapy. Our approach may ultimately be applicable to patient-derived primary cells and cultures of patient tumors to predict effective drugs for individualized care.

Cisplatin-Induced Metabolic Responses Measured with Raman Spectroscopy in Cancer Cells, Spheroids, and Canine-Derived Organoids

Ex vivo assessment of drug response with conventional cell viability assays remains the standard practice for guiding initial therapeutic choices. However, such ensemble approaches fail to capture heterogeneities in treatment response and cannot identify early markers of response. Here, we leverage Raman spectroscopy (RS) as an accurate, low-cost, extraction-free, and label-free approach to track metabolic changes in cancer cells, spheroids, and organoids in response to cisplatin treatment. We identified 12 statistically significant metabolites in cells and 19 metabolites in spheroids and organoids as a function of depth. We show that the cisplatin treatment of 4T1 cells and spheroids results in a shift in metabolite levels; metabolites including nucleic acids such as DNA, 783 cm-1 with p = 0.00021 for cells; p = 0.02173 for spheroids, major amino acids such as threonine, 1338 cm-1 with p = 0.00045 for cells; p = 0.01022 for spheroids, proteins such as amide III, 1248 cm-1 with p = 0.00606 for cells; p = 0.00511 for spheroids serve as early predictors of response. Our RS findings were also applicable to canine-derived organoids, showing spatial variations in metabolic changes as a function of organoid depth in response to cisplatin. Further, the metabolic pathways such as tricarboxylic acid (TCA)/citric acid cycle and glyoxylate and dicarboxylate metabolism that drive drug response showed significant differences based on organoid depth, replicating the heterogeneous treatment response seen in solid tumors where there is a difference from the periphery to the tumor core. Our study showcases the versatility of RS as a predictive tool for treatment response applicable from cells to organotypic cultures, that has the potential to decrease animal burden and readout time for preclinical drug efficacy.

Predicting the onset of preeclampsia by longitudinal monitoring of metabolic changes throughout pregnancy with Raman spectroscopy

Preeclampsia is a life-threatening pregnancy disorder. Current clinical assays cannot predict the onset of preeclampsia until the late 2nd trimester, which often leads to poor maternal and neonatal outcomes. Here we show that Raman spectroscopy combined with machine learning in pregnant patient plasma enables rapid, highly sensitive maternal metabolome screening that predicts preeclampsia as early as the 1st trimester with >82% accuracy. We identified 12, 15 and 17 statistically significant metabolites in the 1st, 2nd and 3rd trimesters, respectively. Metabolic pathway analysis shows multiple pathways corresponding to amino acids, fatty acids, retinol, and sugars are enriched in the preeclamptic cohort relative to a healthy pregnancy. Leveraging Pearson's correlation analysis, we show for the first time with Raman Spectroscopy that metabolites are associated with several clinical factors, including patients' body mass index, gestational age at delivery, history of preeclampsia, and severity of preeclampsia. We also show that protein quantification alone of proinflammatory cytokines and clinically relevant angiogenic markers are inadequate in identifying at-risk patients. Our findings demonstrate that Raman spectroscopy is a powerful tool that may complement current clinical assays in early diagnosis and in the prognosis of the severity of preeclampsia to ultimately enable comprehensive prenatal care for all patients.