The application of artificial intelligence (AI) technology to medical imaging has resulted in great breakthroughs. Given the unique position of ultrasound (US) in prenatal screening, the research on AI in prenatal US has practical significance with its application to prenatal US diagnosis improving work efficiency, providing quantitative assessments, standardizing measurements, improving diagnostic accuracy, and automating image quality control. This review provides an overview of recent studies that have applied AI technology to prenatal US diagnosis and explains the challenges encountered in these applications.

Congenital heart defects (CHDs) are the most common group of major birth defects. Presently there are no clinically used biomarkers for prenatally detecting CHDs. Here, we performed a comprehensive maternal serum proteomics assessment, combined with immunoassays, for the discovery of non-invasive biomarkers for prenatal diagnosis of CHDs. A total of 370 women were included in this study. An isobaric tagging for relative and absolute quantification (iTRAQ) proteomic approach was used first to compare protein profiles in pooled serum collected from women who had CHD-possessing or normal fetuses, and 47 proteins displayed significant differential expressions. Targeted verifications were performed on 11 proteins using multiple reaction monitoring mass spectrometry (MRM-MS), and the resultant candidate biomarkers were then further validated using ELISA analysis. Finally, we identified a biomarker panel composed of 4 cytoskeletal proteins capable of differentiating CHD-pregnancies from normal ones [with an area under the receiver operating characteristic curve (AUC) of 0.938, P < 0.0001]. The discovery of cytoskeletal protein changes in maternal serum not only could help us in prenatal diagnosis of CHDs, but also may shed new light on CHD embryogenesis studies.

Purpose: This study explored the performance of prenatal ultrasonography in the differential diagnosis of cystic biliary atresia (CBA) and choledochal cyst (CC).Methods: Fetuses diagnosed with hepatic hilar cyst in the second trimester were included in this study. A series of prenatal ultrasound examinations were performed in the second and third trimesters. The diameter of the gallbladder (GB) and hepatic cyst were measured, as well as the wall thickness of the GB. The GB-cyst connection, visibility of the right hepatic artery (RHA), and other concomitant abnormalities were carefully evaluated. A neonatal transabdominal ultrasound examination was performed within 1 week after birth, and clinical data were followed up to 6 months after birth.Results: Between January 1, 2016 and January 31, 2020, 53 fetuses diagnosed with hepatic hilar cyst were recruited. Eight were excluded because they were lost to follow-up. Among the 45 cases included in this study, 10 were diagnosed with CBA and 35 with CC after birth. Statistically significant differences were found in GB width, wall thickness, change in GB width, change in cyst length, GB-cyst connection, and RHA visibility between the CBA and CC groups. GB width showed the best diagnostic performance with an area under the curve (AUC) of 0.899. The combination of GB width, GB wall thickness, and GB-cyst connection yielded a comparable AUC of 0.971.Conclusion: The GB should be carefully evaluated in fetuses with hepatic hilar cyst. Prenatal ultrasound findings could provide suggestive parameters for the differential diagnosis of CBA from CC.