Sonographic biometry in the early third trimester: A comparison of parameters to predict macrosomia at birth

A Predictive Model for Large-for-Gestational-Age Infants among Korean Women with Gestational Diabetes Mellitus Using Maternal Characteristics and Fetal Biometric Parameters

Background: With increasing incidence of gestational diabetes mellitus (GDM), newborn infants with perinatal morbidity, including large-for-gestational-age (LGA) or macrosomia, are also increasing. The purpose of this study was to develop a prediction model for LGA infants with GDM mothers. Methods: This was a retrospective case-control study of 660 women with GDM and singleton pregnancies in four tertiary care hospitals from 2006 to 2013 in Korea. Biometric parameters were obtained at diagnoses of GDM and within two weeks before delivery. These biometric data were all transformed retrospectively into Z-scores calculated using a reference. Interval changes of values between the two periods were obtained. Multivariable logistic and stepwise backwards regression analyses were performed to develop the most parsimonious predictive model. The prediction model included pre-pregnancy body mass index (BMI), head circumference (HC), Z-score at 24 + 0 to 30 + 6 weeks’ gestation, and abdominal circumference (AC) Z-score at 34 + 0 to 41 + 6 weeks within 2 weeks before delivery. The developed model was then internally validated. Results: Our model’s predictive performance (area under the curve (AUC): 0.925) was higher than estimated fetal weight (EFW) within two weeks before delivery (AUC: 0.744) and the interval change of EFW Z-score between the two periods (AUC: 0.874). It was internally validated (AUC: 0.916). Conclusions: A clinical model was developed and internally validated to predict fetal overgrowth in Korean women with GDM, which showed a relatively good performance.

Diagnostic value of sonographic fetal anthropometries and anthropometric formulas to identify macrosomia: a meta-analysis

INTRODUCTION

This study evaluated whether sonographic fetal anthropometries and anthropometric formulas can identify macrosomia, with increases in mortality and morbidity rates in infanthood and probably later in life.

EVIDENCE ACQUISITION

Meta-analysis including good-quality studies determined summarized sensitivity, specificity, and positive and negative likelihood ratios (PLR and NLR, respectively) and area under the curve (AUC). PLR and NLR divided informational usability into exclusion and confirmation strategies (10<PLR and NLR<0.1), confirmation strategies only (10<PLR and NLR>0.1), exclusion strategies only (10>PLR and NLR<0.1), or neither exclusion nor confirmation strategies (10>PLR and NLR>0.1). Subgroup and meta-regression analyses were performed.

EVIDENCE SYNTHESIS

Abdominal circumference showed moderately high sensitivity and moderately high specificity (N.=4). However, informational usability classified it as a neither exclusion nor confirmation strategy. Anthropometric formulas showed high specificity (N.=21). However, use of anthropometric formulas showed low sensitivity, and informational usability classified it as a neither exclusion nor confirmation strategy. On the other hand, limiting to Hadlock IV (1985) formula changed this to a confirmation strategy only (N.=7). Hadlock IV (1985) formula versus other formulas may have been a true confounder.

CONCLUSIONS

Abdominal circumference and varying anthropometric formulas are not highly useful for identification of macrosomia. However, Hadlock IV (1985) formula may be useful for secondary screening of macrosomia.

Machine-learning-based automatic identification of fetal abdominal circumference from ultrasound images

OBJECTIVE

Obstetricians mainly use ultrasound imaging for fetal biometric measurements. However, such measurements are cumbersome. Hence, there is urgent need for automatic biometric estimation. Automated analysis of ultrasound images is complicated owing to the patient-specific, operator-dependent, and machine-specific characteristics of such images.

APPROACH

This paper proposes a method for the automatic fetal biometry estimation from 2D ultrasound data through several processes consisting of a specially designed convolutional neural network (CNN) and U-Net for each process. These machine learning techniques take clinicians' decisions, anatomical structures, and the characteristics of ultrasound images into account. The proposed method is divided into three steps: initial abdominal circumference (AC) estimation, AC measurement, and plane acceptance checking.

MAIN RESULTS

A CNN is used to classify ultrasound images (stomach bubble, amniotic fluid, and umbilical vein), and a Hough transform is used to obtain an initial estimate of the AC. These data are applied to other CNNs to estimate the spine position and bone regions. Then, the obtained information is used to determine the final AC. After determining the AC, a U-Net and a classification CNN are used to check whether the image is suitable for AC measurement. Finally, the efficacy of the proposed method is validated by clinical data.

SIGNIFICANCE

Our method achieved a Dice similarity metric of [Formula: see text] for AC measurement and an accuracy of 87.10% for our acceptance check of the fetal abdominal standard plane.

Lithium exposure during pregnancy: outcomes for women who attended a specialist antenatal clinic

INTRODUCTION

Lithium treatment in pregnancy represents a significant dilemma for women and treating health professionals alike. The complexity of risk-benefit analysis is impacted by limited information.

METHODS

A cohort study of 33 women with severe mental illness, who were prescribed lithium at any time during the pregnancy, and gave birth between December 2007 and January 2015 at a specialist antenatal clinic in Western Australia. A descriptive comparison for women who continued lithium throughout pregnancy, and those who ceased on discovery of pregnancy was undertaken examining demographic, obstetric, neonatal and psychiatric variables.

RESULTS

Women who were prescribed lithium, irrespective of whether they continued or discontinued the medication represented a high risk group obstetrically, with high rates of smoking overall (33%) medical comorbidities (54%) and antenatal complications (88%). Preconception counseling occurred in 33% of the cohort but increased the likelihood of continuing lithium in pregnancy (p = .007). Compared to those who ceased lithium, women who remained on lithium through the pregnancy had increased rates of fetal ultrasound abnormalities such as abdominal circumference >90th % (p = .005). Psychiatric relapses through the antenatal and immediate postpartum period appeared to be due to a combination of factors.

DISCUSSION

Pregnant women with severe mood disorders treated with lithium are a vulnerable, high-risk obstetric population who would benefit from preconception counseling, regular antenatal care in a tertiary center, delivery with neonatal pediatric support and experienced psychiatric management.

Serum MicroRNAs as Diagnostic Biomarkers for Macrosomia

BACKGROUND

Macrosomia is defined as an infant's birth weight of more than 4000 g. Although microRNAs (miRNAs) have been implicated in the pathogenesis of various diseases, the associations between serum miRNAs and macrosomia have been rarely reported.

METHODOLOGY

We used the Taqman Low Density Array followed by quantitative reverse transcriptase polymerase chain reaction assays to screen for miRNAs associated with macrosomia using serum samples collected 1 week before delivery.

RESULTS

Profiling results showed that 1 miRNA was significantly upregulated and 10 miRNAs were significantly downregulated in serum samples of macrosomia (ΔΔCt > 3-fold). The expression levels of miR-21 were significantly decreased in macrosomia as compared to the controls in the third trimester. Receiver operating characteristic (ROC) curve analyses showed that the area under the ROC curve for miR-21 was 67.7% (sensitivity = 66.7% and specificity = 70.0%).

CONCLUSIONS

miR-21 in maternal serum is differentially expressed between macrosomia and controls, and miR-21 could be used as a candidate biomarker to predict macrosomia.