(1) Background: Artificial Intelligence (AI) is a modern tool with numerous applications in the medical field. The case series reported here aimed to investigate the diagnostic performance of the fetal intelligent navigation echocardiography (FINE) method applied for the first time in the prenatal identification of atrioventricular septal defects (AVSD). This congenital heart disease (CHD) is associated with extracardiac anomalies and chromosomal abnormalities. Therefore, an early diagnosis is essential to advise parents and make adequate treatment decisions. (2) Methods: Four fetuses diagnosed with AVSD via two-dimensional (2D) ultrasound examination in the second trimester were enrolled. In all cases, the parents chose to terminate the pregnancy. Since the diagnosis of AVSD with 2D ultrasound may be missed, one or more four-dimensional (4D) spatiotemporal image correlation (STIC) volume datasets were obtained from a four-chamber view. The manual navigation enabled by the software is time-consuming and highly operator-dependent. (3) Results: FINE was applied to these volumes and nine standard fetal echocardiographic views were generated and optimized automatically, using the assistance of the virtual intelligent sonographer (VIS). Here, 100% of the four-chamber views, and after the VISA System application the five-chamber views, of the diagnostic plane showed the atrioventricular septal defect and a common AV valve. The autopsies of the fetuses confirmed the ultrasound results. (4) Conclusions: By applying intelligent navigation technology to the STIC volume datasets, 100% of the AVSD diagnoses were detected.
Keywords: atrioventricular septal defects; congenital heart disease; fetal intelligent navigation echocardiography; spatiotemporal image correlation; virtual intelligent sonographer.