Mean Fetal Volume Research Articles

Automatic placental and fetal volume estimation by a convolutional neural network

IntroductionWe aimed to develop an artificial intelligence (AI) deep learning algorithm to efficiently estimate placental and fetal volumes from magnetic resonance (MR) scans. MethodsManually annotated images from an MRI sequence was used as input to the neural network DenseVNet. We included data from 193 normal pregnancies at gestational week 27 and 37. The data were split into 163 scans for training, 10 scans for validation and 20 scans for testing. The neural network segmentations were compared to the manual annotation (ground truth) using the Dice Score Coefficient (DSC). ResultsThe mean ground truth placental volume at gestational week 27 and 37 was 571 cm3 (Standard Deviation (SD) 293 cm3) and 853 cm3 (SD 186 cm3), respectively. Mean fetal volume was 979 cm3 (SD 117 cm3) and 2715 cm3 (SD 360 cm3). The best fitting neural network model was attained at 22,000 training iterations with mean DSC 0.925 (SD 0.041). The neural network estimated mean placental volumes at gestational week 27–870 cm3 (SD 202 cm3) (DSC 0.887 (SD 0.034), and to 950 cm3 (SD 316 cm3) at gestational week 37 (DSC 0.896 (SD 0.030)). Mean fetal volumes were 1292 cm3 (SD 191 cm3) and 2712 cm3 (SD 540 cm3), with mean DSC of 0.952 (SD 0.008) and 0.970 (SD 0.040). The time spent for volume estimation was reduced from 60 to 90 min by manual annotation, to less than 10 s by the neural network. ConclusionThe correctness of neural network volume estimation is comparable to human performance; the efficiency is substantially improved.

Three-dimensional ultrasound first trimester fetal volume measurement and its relation to pregnancy outcome.

To determine whether fetal volume (FV) measured by three-dimensional (3D) ultrasound was able to detect fetuses at risk of low birth weight (primary outcome) and/or preterm labor (secondary outcome). One hundred pregnant women carrying a singleton living pregnancy who were sure of dates, and had a dating scan, with gestational age between 11 weeks and 13 weeks+6 days coming for routine first trimester nuchal translucency (NT) were examined by both two-dimensional (2D) and 3D ultrasound (Vocal System) for crown-rump length (CRL) and FV then followed up regularly every 4 weeks until 28 weeks then biweekly until 36 weeks then weekly until delivery both clinically and by ultrasound biometry. Eighty-seven cases had a normal outcome, while the remaining 13 cases had either preterm labor (four cases) or low-birth weight (nine cases). FV positively correlated with CRL (P=0.026), gestational age in weeks (P=0.002), neonatal body weight in grams (P=0.018) and neonatal body length at birth (P=0.04). A mean FV of 8.3 mm3 was association with neonatal complications (P=0.045). A cut-off point of 9 mm3 for FV was associated with 100% sensitivity for detection of the date of birth, while a cut-off point of 9.15 mm3 for FV was associated 100% sensitivity for detection of neonatal birth weight. 3D assessment of FV in the first trimester provides an accurate method for predicting pregnancy outcome namely low birth weight and neonatal complications, however, it is a better positive predictor than a negative one.

P30.07: Predicting gastroschisis perinatal outcome from antenatal characteristics

Objectives: The objective was to estimate a reference range of 3D embryo volume using VOCAL for pregnancies between 7 & 11weeks. Methods: This is a cross-sectional study including 60 singleton pregnancies with an E6 (GE, USA)with an endocavitary volumetric 4–9 MHz probe, using VOCAL with a manual tracer at 30◦, the outer aspect of the embryo was drawn manually on every image displayed sequentially in this plane. At the end of the 180◦ rotation, the built in software calculated the volume automatically. Results: The mean CRL was 23.7 ± 9.0 mm (range 9 to 38 mm). Mean fetal volume was 1.7 ± 1.2 cm3 (range 0.35 to 4.0 cm3). There was a strong positive correlation between fetal volume and menstrual age, gestational age and CRL (r = 0.919, 0.938 and 0.941, resp.). Power regression model produced an R2 value of 0.880 with a regression equation (y = 8.657 * ×0.150). The predicted gestational age fitted well with that predicted from CRL with a minimal difference and a P of 0.234. Conclusions: This pilot study demonstrated a strong correlation between fetal volume estimation and crown-rump length measurement for the prediction of gestational age (r = 0.950). Embryonic/fetal volume is a good predictor of gestational age with a power regression equation (y = 8.657 * ×0.150) for the period from 7 weeks to 10 weeks + 6 days.

P30.05: 3D ultrasonography using VOCAL for estimation of reference range of 1st trimester fetal volume

Objectives: The objective was to estimate a reference range of 3D embryo volume using VOCAL for pregnancies between 7 & 11weeks. Methods: This is a cross-sectional study including 60 singleton pregnancies with an E6 (GE, USA)with an endocavitary volumetric 4–9 MHz probe, using VOCAL with a manual tracer at 30◦, the outer aspect of the embryo was drawn manually on every image displayed sequentially in this plane. At the end of the 180◦ rotation, the built in software calculated the volume automatically. Results: The mean CRL was 23.7 ± 9.0 mm (range 9 to 38 mm). Mean fetal volume was 1.7 ± 1.2 cm3 (range 0.35 to 4.0 cm3). There was a strong positive correlation between fetal volume and menstrual age, gestational age and CRL (r = 0.919, 0.938 and 0.941, resp.). Power regression model produced an R2 value of 0.880 with a regression equation (y = 8.657 * ×0.150). The predicted gestational age fitted well with that predicted from CRL with a minimal difference and a P of 0.234. Conclusions: This pilot study demonstrated a strong correlation between fetal volume estimation and crown-rump length measurement for the prediction of gestational age (r = 0.950). Embryonic/fetal volume is a good predictor of gestational age with a power regression equation (y = 8.657 * ×0.150) for the period from 7 weeks to 10 weeks + 6 days.

P30.06: Intrapartum translabial sonography: a prospective analysis of various methods to predict the mode of delivery in prolonged second stage of labor

Objectives: The objective was to estimate a reference range of 3D embryo volume using VOCAL for pregnancies between 7 & 11weeks. Methods: This is a cross-sectional study including 60 singleton pregnancies with an E6 (GE, USA)with an endocavitary volumetric 4–9 MHz probe, using VOCAL with a manual tracer at 30◦, the outer aspect of the embryo was drawn manually on every image displayed sequentially in this plane. At the end of the 180◦ rotation, the built in software calculated the volume automatically. Results: The mean CRL was 23.7 ± 9.0 mm (range 9 to 38 mm). Mean fetal volume was 1.7 ± 1.2 cm3 (range 0.35 to 4.0 cm3). There was a strong positive correlation between fetal volume and menstrual age, gestational age and CRL (r = 0.919, 0.938 and 0.941, resp.). Power regression model produced an R2 value of 0.880 with a regression equation (y = 8.657 * ×0.150). The predicted gestational age fitted well with that predicted from CRL with a minimal difference and a P of 0.234. Conclusions: This pilot study demonstrated a strong correlation between fetal volume estimation and crown-rump length measurement for the prediction of gestational age (r = 0.950). Embryonic/fetal volume is a good predictor of gestational age with a power regression equation (y = 8.657 * ×0.150) for the period from 7 weeks to 10 weeks + 6 days.

OC16.05: First trimester human yolk sac and fetal volumetry in virtual reality

Objective: The aim of this study was to evaluate real-time three dimensional (3D) determined volumes of the yolk sac and fetal body, including the limbs, in first trimester pregnancies using a virtual reality (VR) application, and to analyse relations with gestational age (GA) and crown-rump length (CRL). Methods: We analyzed 180 3D first trimester ultrasound scans of 42 spontaneous and IVF/ICSI pregnancies. Only scans with high quality images were included and transferred to a Barco ISpace VR system and visualized as 3D ‘holograms’ with V-Scope volume rendering software. An implemented segmentation algorithm calculated the yolk sac volume (YSV) and fetal volume (FV) semiautomatically (A). The logarithmically transformed outcomes were analyzed using repeated measurements ANOVA. Interobserver agreement was established by calculating intraclass correlation coefficients (ICC). Results: In total 118 YSV and 88 FV were measured with GA between 5 + 5 and 12 + 6 weeks. YSV ranged from 33 mm3 to 424 mm3 and FV ranged from 14 mm3 to 29877 mm3. Reference charts were constructed according to GA and CRL (B). ANOVA calculations showed that when the CRL doubles, the mean FV increases 6.5 fold and when the GA doubles the mean FV increases approximately 500 fold (p<0.001). Furthermore, it was found that for each day, the YSV increases on average by 4% and when the CRL doubles, the YSV increases with 67% (p<0.001). Interobserver agreement was excellent with an ICC of 0.99. Conclusions: We constructed reference charts for growth of YSV and FV in the first trimester using a VR system. This approach allows us to obtain information about the size of the fetus using all dimensions. These new data can be used to differentiate between normal and abnormal human development.

Vascular volumes in isolated perfused guinea pig placenta.

Maternal and fetal vascular volumes were determined in 13 isolated artificially perfused guinea pig placentas by measuring mean transit times of an intravascular indicator (Evan's blue dye) at constant flow rates. When both maternal and fetal flow rates were 3.1 ml/min, the average maternal volume of the placenta was 1.85 +/- 0.54 (SD) ml, and the mean fetal volume was 0.92 +/- 0.2 (SD) ml. If calculated maternal volumes were corrected for the myometrial vascular volume and for the interlobium volume, the remaining volume attributable largely to the labyrinth averaged approximately 1.2 ml. When flow rates were changed on the fetal or the maternal side between 0 and 6.2 ml/min, vascular volumes also changed. For example, volumes increased directly with flow rates on that side of the placenta in which the flow change was introduced; in most cases, it decreased on the opposite side where the flow rates had not been altered. This intraplacental volume shift may be regarded as the basic event for the sluice flow phenomenon in placentas.