Predictive Assessment of Early Pregnancy Outcomes

Early pregnancy loss remains a common clinical challenge, prompting extensive research into methods for reliable prediction during the first trimester. Advances in transvaginal ultrasound have enabled detailed measurement of structures such as the gestational sac, yolk sac, crown-rump length and embryonic heart rate, while parallel progress in assay technologies has refined serial monitoring of hormones including β-human chorionic gonadotropin, progesterone and oestradiol. Statistical modelling—ranging from logistic regression to dynamic hazard models—and emerging machine-learning tools seek to integrate these parameters with maternal demographics and clinical findings such as vaginal bleeding. By identifying high-risk pregnancies before overt symptoms, predictive assessment aims to inform patient counselling, personalise surveillance schedules and optimise allocation of clinical resources. The growing convergence of sonographic, biochemical and algorithmic approaches offers a multifaceted framework for early risk stratification and supports the global imperative to improve outcomes in assisted and spontaneous conceptions.

Research from Nature Portfolio

In a prospective cohort of first-trimester pregnancies, researchers serially measured gestational sac diameter, yolk sac diameter, crown-rump length and fetal heart rate from six to ten weeks and constructed reference charts of their trajectories in ongoing versus failing pregnancies. Abnormal sac sizes were the earliest indicators of loss, with small gestational sacs and enlarged yolk sacs deviating from normal as early as six weeks; subsequent deviations in heart rate and embryo length emerged at seven to eight weeks. Validation of logistic models based on these dynamic measurements demonstrated strong discrimination for first-trimester failure, underscoring their utility for early patient counselling and cost-effective care planning.

Predictive Assessment of Early Pregnancy Outcomes publication trend

The graph below shows the total number of articles in predictive assessment of early pregnancy outcomes across all publications each year (not limited to Nature Index journals).

Technical terms

Crown-rump length (CRL): the maximum length of the embryo or fetus from head to rump measured by ultrasound, used to assess gestational age and growth trajectories.

Gestational sac (GS) diameter: the average internal width of the fluid-filled gestational sac as seen on ultrasound, indicative of early embryonic development.

Yolk sac (YS) diameter: the size of the secondary sac within the gestational sac that provides early nutritional support to the embryo.

Fetal heart rate (FHR): the frequency of the embryonic heartbeat measured via ultrasound Doppler, reflecting viability and development.

β-Human chorionic gonadotropin (β-HCG): a hormone produced by the placental trophoblast, widely used as a biochemical marker of early pregnancy progression.

Logistic regression model: a statistical tool that estimates the probability of a binary outcome (e.g. ongoing pregnancy versus loss) based on multiple predictor variables.

Convolutional neural network (CNN): a deep learning architecture designed to process image data by learning hierarchical feature patterns, applied here to ultrasound image analysis.

References

1. A simple scoring system for the prediction of early pregnancy loss developed by following 13,977 infertile patients after in vitro fertilization. European Journal of Medical Research (2023).
2. Early pregnancy ultrasound measurements and prediction of first trimester pregnancy loss: A logistic model. Scientific Reports (2020).
3. A Novel Approach to Predicting Early Pregnancy Outcomes Dynamically in a Prospective Cohort Using Repeated Ultrasound and Serum Biomarkers. Reproductive Sciences (2023).
4. Automated prediction of early spontaneous miscarriage based on the analyzing ultrasonographic gestational sac imaging by the convolutional neural network: a case-control and cohort study. BMC Pregnancy and Childbirth (2022).