This study addresses congenital heart disease in fetuses, a specific birth defect, by collaborating with multicenter medical institutions across China. It prospectively collects echocardiographic screening, diagnostic images, and structured parameter data from both normal and abnormal fetuses. Cases are followed up, and artificial intelligence (AI) models are developed to extract fetal heart echocardiographic image features and analyze the data. This process allows for the analysis of the AI models' accuracy, sensitivity, and specificity, and verifies the application value of AI models in fetal heart echocardiographic screening and diagnosis.

Inclusion Criteria:

• Single or multiple pregnancy;
• Gestational age 16-40 weeks;
• Fetal echocardiographic images are clear and complete;
• Clinical data are complete and detailed;
• Sign informed consent and be able to cooperate with follow-up

Exclusion Criteria:

• Fetal echocardiographic images are incomplete or of poor quality that cannot be used for research;
• Refusal or loss of follow-up

The study population was from 20 multi-center medical institutions in China, and pregnant women who underwent fetal echocardiography between 16 and 40 weeks of gestation.