FetalCLIP introduces a specialized foundation model for fetal ultrasound image analysis that overcomes the limitations of existing approaches by effectively handling the complexity of prenatal imaging.

• Adapts the CLIP architecture specifically for fetal ultrasound interpretation
• Performs multiple critical tasks including classification, gestational age estimation, and congenital heart defect detection
• Demonstrates superior performance despite the scarcity of paired multimodal data
• Enables more accurate fetal structure segmentation

This research represents a significant advancement in prenatal diagnostics, potentially improving clinical workflows and enabling more accurate, efficient fetal assessments in medical settings where specialized expertise may be limited.

FetalCLIP: A Visual-Language Foundation Model for Fetal Ultrasound Image Analysis