A deep learning approach to identify the fetal head position using transperineal ultrasound during labor.

OBJECTIVES: To develop a deep learning (DL)-model using convolutional neural networks (CNN) to automatically identify the fetal head position at transperineal ultrasound in the second stage of labor. MATERIAL AND METHODS: Prospective, multicenter study including singleton, term, cephalic pregnancies in the second stage of labor. We assessed the fetal head position using transabdominal ultrasound and subsequently, obtained an image of the fetal head on the axial plane using transperineal ultrasound and labeled it according to the transabdominal ultrasound findings. The ultrasound images were randomly allocated into the three datasets containing a similar proportion of images of each subtype of fetal head position (occiput anterior, posterior, right and left transverse): the training dataset included 70 %, the validation dataset 15 %, and the testing dataset 15 % of the acquired images. The pre-trained ResNet18 model was employed as a foundational framework for feature extraction and classification. CNN1 was trained to differentiate between occiput anterior (OA) and non-OA positions, CNN2 classified fetal head malpositions into occiput posterior (OP) or occiput transverse (OT) position, and CNN3 classified the remaining images as right or left OT. The DL-model was constructed using three convolutional neural networks (CNN) working simultaneously for the classification of fetal head positions. The performance of the algorithm was evaluated in terms of accuracy, sensitivity, specificity, F1-score and Cohen's kappa. RESULTS: Between February 2018 and May 2023, 2154 transperineal images were included from eligible participants across 16 collaborating centers. The overall performance of the model for the classification of the fetal head position in the axial plane at transperineal ultrasound was excellent, with an of 94.5 % (95 % CI 92.0--97.0), a sensitivity of 95.6 % (95 % CI 96.8-100.0), a specificity of 91.2 % (95 % CI 87.3-95.1), a F1-score of 0.92 and a Cohen's kappa of 0.90. The best performance was achieved by the CNN1 - OA position vs fetal head malpositions - with an accuracy of 98.3 % (95 % CI 96.9-99.7), followed by CNN2 - OP vs OT positions - with an accuracy of 93.9 % (95 % CI 89.6-98.2), and finally, CNN3 - right vs left OT position - with an accuracy of 91.3 % (95 % CI 83.5-99.1). CONCLUSIONS: We have developed a DL-model capable of assessing fetal head position using transperineal ultrasound during the second stage of labor with an excellent overall accuracy. Future studies should validate our DL model using larger datasets and real-time patients before introducing it into routine clinical practice.

Four cases of cervical varices without placenta praevia: Presentation, diagnosis, managements, and literature review.

Cervical varices are a rare condition characterized by recurrent antepartum hemorrhage and less than 20 cases were reported in the literature. It is usually associated with placenta previa. We herein describe four cases of cervical varices without placenta previa. Meticulous speculum examination, ultrasonography with Doppler and colposcopy are essential for establishing the diagnosis and assessing the extent of the cervical varix. We propose to classify it as the apparent external os type or ultrasonography-based endocervical type. Most cases presented in the literature were delivered by cesarean section. Nevertheless, one of our cases was a successful vaginal delivery. Our case illustrates that vaginal delivery is possible in isolated cervical varices. More case reports are needed to have a better understanding of this rare entity.

Influence of Maternal Infection and Pregnancy Complications on Cord Blood Telomere Length.

BACKGROUND: Exposure to suboptimal intrauterine environment might induce structural and functional changes that can affect neonatal health. Telomere length as an important indicator of cellular health has been associated with increased risk for disease development. OBJECTIVES: This study was aimed to examine the independent and combined effects of maternal, obstetric, and foetal factors on cord blood telomere length (TL). METHODS: Pregnant women at the gestational age of 20th to 24th week who attended the antenatal clinic of a major local hospital in Hong Kong were recruited. Participants were asked to complete a questionnaire on demographics, health-related quality of life, and history of risk behaviors. Medical history including pregnancy complications and neonatal outcomes was obtained from electronic medical records of both mother and neonate. Umbilical cord blood was collected at delivery for TL determination. RESULTS: A total of 753 pregnant women (average age: 32.18 ± 4.51 years) were recruited. The prevalence of maternal infection, anaemia, and hypertension during pregnancy was 30.8%, 30.0%, and 6.0%, respectively. The adjusted regression model displayed that maternal infection was negatively associated with cord blood TL (ß = -0.18, p = 0.026). This association became even stronger in the presence of antenatal anaemia, hypertension, delivery complications, or neonatal jaundice (ß = -0.25 to -0.45). CONCLUSIONS: This study consolidates evidence on the impact of adverse intrauterine environment at the cellular level. Maternal infection was significantly associated with shorter cord blood TL in a unique manner such that its presence may critically determine the susceptibility of telomere to other factors.

Asynclitism in the second stage of labor: prevalence, associations, and outcome.

BACKGROUND: Fetal head asynclitism may affect labor progress, increase the need for obstetrical intervention, and even be associated with difficult or failed instrumental delivery. However, there is limited evidence on the true prevalence and associations of asynclitism when diagnosed by transperineal ultrasound in the second stage of labor. OBJECTIVE: This study aimed to examine the prevalence and outcome of asynclitism in the second stage of labor in a regional hospital in Hong Kong. STUDY DESIGN: This is a prospective cohort study involving 92 term nulliparous women with singleton pregnancy in the second stage of labor, recruited from December 2019 to December 2020. Transperineal ultrasound was performed and asynclitism was diagnosed if there was asymmetry of intracranial structures on the transverse plane. To assess the fetal head station, the head perineum distance was measured at rest and on pushing, and the mode of delivery was recorded. Relationship between asynclitism and fetal head position, and between head perineum distance and the mode of delivery, were evaluated with chi-squared tests and Mann-Whitney U tests. RESULTS: The prevalence of asynclitism was 15% (14 of 92), of which 12 were anterior asynclitism and 2 were posterior asynclitism. The prevalence of asynclitism was less common in occiput anterior compared with nonocciput anterior position (6.7% vs 53%, P<.01). Women with asynclitism were associated with smaller delta head perineum distance (head perineum distance at rest minus that at pushing) than women without asynclitism (median [interquartile range], 0.68 cm [0.85 cm] vs 0.91 cm [0.71 cm]; P=.01). Eventually, there was a trend of more operative deliveries in women with asynclitism (43%, or 6 of 14 women) than in women without asynclitism (27%, or 21 of 78 women), although this difference was not statistically significant (P=.22). CONCLUSION: The prevalence of asynclitism at transperineal ultrasound was rather common in nulliparous women at second stage of labor and seemed more commonly associated with nonocciput anterior position.