Using deep learning to predict the outcome of live birth from more than 10,000 embryo data

Background Recently, the combination of deep learning and time-lapse imaging provides an objective, standard and scientific solution for embryo selection. However, the reported studies were based on blastocyst formation or clinical pregnancy as the end point. To the best of our knowledge, there is no predictive model that uses the outcome of live birth as the predictive end point. Can a deep learning model predict the probability of live birth from time-lapse system? Methods This study retrospectively analyzed the time-lapse data and live birth outcomes of embryos samples from January 2018 to November 2019. We used the SGD optimizer with an initial learning rate of 0.025 and cosine learning rate reduction strategy. The network is randomly initialized and trained for 200 epochs from scratch. The model is quantitively evaluated over a hold-out test and a 5-fold cross-validation by the average area under the curve (AUC) of the receiver operating characteristic (ROC) curve. Results The deep learning model was able to predict live birth outcomes from time-lapse images with an AUC of 0.968 in 5-fold stratified cross-validation. Conclusions This research reported a deep learning model that predicts the live birth outcome of a single blastocyst transfer. This efficient model for predicting the outcome of live births can automatically analyze the time-lapse images of the patient’s embryos without the need for manual embryo annotation and evaluation, and then give a live birth prediction score for each embryo, and sort the embryos by the predicted value.

Determinants of Stillbirth From Two Observational Studies Investigating Deliveries in Kano, Nigeria

Background:Stillbirths are a poignant representation of global inequality. Nigeria is documented to have the second highest rate; yet, the reporting system is inadequate in most Nigerian healthcare facilities. The aim was to identify the determinants of stillbirth among deliveries in the Murtala Muhammad Specialist Hospital (MMSH), Kano, Nigeria.Methods:Two study designs were used: a case-control study (S1) and a prospective cohort study (S2). Both studies were carried out at the MMSH. For S1, stillbirths were retrospectively matched to a livebirth by time (target of 24 hours' time variation) to establish a case-control study with a 1:1 ratio. Eligibility into S2 included all mothers who were presented at the MMSH in labour regardless of birth outcome. Both were based on recruitment durations, not sample sizes (3 months and 2 months, respectively, 2017–2018). The demographic and clinical data were collected through paper-based questionnaires. Univariable logistic regression was used. Multivariable logistic regression was used to explore relationships between area type and other specific factors.Findings:Stillbirth incidence in S2 was 180/1,000 births. Stillbirth was associated with the following factors; no maternal education, previous stillbirth(s), prematurity, living in both semi-rural and rural settings, and having extended time periods between rupture of membranes and delivery. Findings of the multivariable analysis (S1 and S2) indicated that the odds of stillbirth, for those living in a rural area, were further exacerbated in those mothers who had no education, lived in a shack, or had any maternal disease.Interpretation:This research identifies the gravity of this situation in this area and highlights the need for action. Further understanding of some of the findings and exploration into associations are required to inform intervention development.Funding:This collaboration was partially supported by funding from Health and Care Research Wales.

Gardnerella Vaginalis Alters Cervicovaginal Epithelial Cell Function Through Epithelial Cell-type Specific Immune Responses

Background: The cervicovaginal (CV) microbiome is highly associated with vaginal health and disease in both pregnant and non-pregnant individuals. An overabundance of Gardnerella vaginalis in the CV space is commonly associated with adverse reproductive outcomes including bacterial vaginosis (BV), sexually transmitted diseases and preterm birth while the presence of Lactobacillus spp is often associated with reproductive health. While host-microbial interactions are hypothesized to contribute to CV health and disease, the mechanisms by which these interactions regulate CV epithelial function remain largely unknown. Results: Using an in vitro co-culture model, we assessed the effects of Lactobacillus crispatus and G. vaginalis on the CV epithelial barrier, the immune mediators that could be contributing to decreased barrier integrity and the immune signaling pathways regulating the immune response. G. vaginalis, but not L. crispatus, significantly increased epithelial cell death and decreased epithelial barrier integrity in an epithelial cell-specific manner. A G. vaginalis-mediated epithelial immune response including NFkB activation and proinflammatory cytokine release was initiated partially through TLR2 dependent signaling pathways. Additionally, investigation of the cytokine immune profile in human CV fluid showed distinctive clustering of cytokines by G. vaginalis abundance and birth outcome. Conclusions: The results of this study show both microbe- and epithelial cell-type specific effects on CV epithelial function. Altered epithelial barrier function through cell death and immune mediated mechanisms by G. vaginalis, but not L. crispatus, indicates that host epithelial cells respond to bacteria-associated signals, resulting in altered epithelial function and ultimately CV disease. Additionally, distinct immune signatures associated with G. vaginalis or birth outcome provide further evidence that host-microbial interactions may contribute significantly to the biological mechanisms regulating reproductive outcomes.

Evaluation of Risk Factors for Preterm Birth Outcome in Gujarat, India

Background: Preterm birth (PTB) is a leading cause of neonatal survival complications, mortality, and morbidity worldwide. In India 35% of all neonatal deaths are due to PTB with 36th global ranking, hence, India's healthcare sector has been working towards reducing the rate of PTB effectively. Objective: This study aimed to assess the risk factors such as environmental and pathophysiological causes associated with preterm birth in the population of Gujarat, India. Materials and Methods: In this study, multivariate random sampling was performed and systematically 200 pregnant mothers [PTB <37 weeks (N=100), Full-term >37 weeks (N=100)] were chosen after excluding mothers with vaginal infection, multiple gestations, fetal anomalies, non-cephalic presentation, cesarean delivery, and pregnancy with Mullerian anomalies. Statistical analysis was performed by Chi-square test, and variables with p-value <0.05 were considered statistically significant. Results: Out of all the variables, maternal age below 20 years, extreme BMI, high blood pressure during pregnancy, maternal health complications, medication and doctor's consultation were highly significant (p < 0.0001). Furthermore, variables like type of area, diet, education, Hb levels below 9 g/dL and above 13 g/dL and blood-group of the mother were also significantly associated with PTB outcome (p < 0.05). Among the sub-categories of PTB (extreme-, very-, moderate/late- PTB), maternal age, Hb level and the past obstetric outcome showed very high significance (p < 0.0001). Conclusion: For the prediction of birth outcome, mother's internal physiological and lifestyle factors need to be taken into consideration, and mothers at risk priorly can be screen out, followed by proper healthcare assistance to decrease the preterm birth rate and its consequences. Key words: Preterm birth, maternal age, blood pressure, past obstetric outcome, risk factors

Determinants of adverse birth outcomes among women delivered in public hospitals of Ethiopia, 2020

Background Adverse birth outcome is a common health problem consisting of several health effects involving pregnancy and the newborn infant. Infants with one or more adverse birth outcomes are at greater risk for mortality and a variety of health and developmental problems. Factors such as the age of the mother, antepartum hemorrhage, history of abortion, gestational age, anemia, and maternal undernutrition have predisposed the mother to adverse birth outcome. For appropriate prevention of the adverse birth outcomes, data pertaining to determinants of adverse birth outcomes are important. Therefore, this study was aimed to assess the determinants of adverse birth outcomes among women who give birth in public hospitals of western Ethiopia. Methods An institutional-based unmatched prospective case-control study was conducted from February 15 to April 15, 2020, in selected public hospitals of western Ethiopia. From mothers who gave birth in public hospitals of Wollega zones, 165 cases and 330 controls were selected. Mothers with adverse birth outcomes were cases and mothers without adverse birth were controls. Data was collected by structured interviewer-administered questionnaires. In addition to the interview, the data collectors abstracted clinical data by reviewing the mother and the babies’ medical records. The collected data were entered into Epi info version 7 and exported to SPSS version 21 for analysis. Finally, multivariable logistic regression was used to identify determinants of adverse birth outcomes at P-value < 0.05. Results A total of 495 mothers (165 cases and 330 controls) were included in the study with a mean age of 28.48 + 5.908. Low ANC visit (AOR = 3.92: 95% CI; 1.86, 8.2), premature rupture of membrane (AOR = 2.83: 95% CI; 1.72,4.64), being Anemic (AOR = 2: 95% CI; 1.16,3.44), pregnancy induced-hypertension (AOR = 2.3:95% CI; 1.4,3.85), not getting dietary supplementation (AOR = 2.47:95% CI; 1.6,3.82), and physical abuse (AOR = 2.13: 95% CI; 1.05,4.32) were significantly associated with the development of the adverse birth outcome. Conclusion Low antenatal care visit, being anemic, premature rupture of membrane, pregnancy-induced hypertension, not getting dietary supplementation, and physical abuse were determinants of adverse birth outcomes. The clinicians should play a pivotal role to improve antenatal care follow up, counsel, and supplement recommended diets and minimize violence and abuse during pregnancy.

The effect of secretor status and the vaginal microbiome on birth outcome

Mutations in the FUT2 gene that result in a lack of expression of histo-blood group antigens on secreted glycoproteins may shape the vaginal microbiota with consequences for birth outcome. To test this, we analysed the relationship between secretor status, vaginal microbiota and gestational length in an ethnically diverse cohort of 313 pregnant women, including 91 who delivered prematurely. Lactobacillus species were found to co-occur less often with other microbial taxa in non-secretors. Moreover, non-secretors with Lactobacillus spp. depleted vaginal microbiota in early pregnancy had significantly shorter gestational length than Lactobacillus spp. dominated non-secretors (mean of 245.5 (SD=44.5) versus 265.9 (23.6)); p=0.045), but not compared to Lactobacillus spp. dominated (261.8 (27.5)) and depleted (264.3 days (21.2)) secretors. In identifying a relationship between blood-group antigen expression and vaginal microbiota-host interactions, our results point towards stratification by secretor status as an important factor for considering preterm birth risk and prevention.

A non-optimal cervicovaginal microbiota in pregnancy is associated with a distinct metabolomic signature among non-Hispanic Black individuals

Biomechanical and molecular processes of premature cervical remodeling preceding spontaneous preterm birth (sPTB) likely result from interactions between the cervicovaginal microbiota and host immune responses. A non-optimal cervicovaginal microbiota confers increased risk of sPTB. The cervicovaginal space is metabolically active in pregancy; microbiota can produce, modify, and degrade metabolites within this ecosystem. We establish that cervicovaginal metabolomic output clusters by microbial community in pregnancy among Black individuals, revealing increased metabolism within the amino acid and dipeptide pathways as hallmarks of a non-optimal microbiota. Few differences were detected in metabolomic profiles when stratified by birth outcome. The study raises the possibility that metabolites could distinguish women with greater risk of sPTB among those with similar cervicovaginal microbiota, and that metabolites within the amino acid and carbohydrate pathways may play a role in this distinction.

Prenatal Lead and Depression Exposures Jointly Influence Birth Outcomes and NR3C1 DNA Methylation

Many gestational exposures influence birth outcomes, yet the joint contribution of toxicant and psychosocial factors is understudied. Moreover, associated gestational epigenetic mechanisms are unknown. Lead (Pb) and depression independently influence birth outcomes and offspring NR3C1 (glucocorticoid receptor) DNA methylation. We hypothesized that gestational Pb and depression would jointly influence birth outcomes and NR3C1 methylation. Pregnancy exposure information, DNA methylation, and birth outcome data were collected prospectively from n = 272 mother–infant pairs. Factor analysis was used to reduce the dimensionality of NR3C1. Multivariable linear regressions tested for interaction effects between gestational Pb and depression exposures with birth outcomes and NR3C1. Interaction effects indicated that higher levels of Pb and depression jointly contributed to earlier gestations, smaller infant size at birth, and asymmetric fetal growth. Pb and depression were also jointly associated with the two primary factor scores explaining the most variability in NR3C1 methylation; NR3C1 scores were associated with some infant outcomes, including gestational age and asymmetric fetal growth. Pb and depression can cumulatively influence birth outcomes and epigenetic mechanisms, which may lay the foundation for later health risk. As toxicants and social adversities commonly co-occur, research should consider the life course consequences of these interconnected exposures.