The SwiFT (Swin 4D fMRI Transformer) model represents a breakthrough in early detection of neurodevelopmental issues by analyzing neonatal brain scans.

• Leverages advanced transformer architecture optimized for 4D functional MRI data
• Predicts Bayley-III developmental scores from brain scans in the first months of life
• Enables earlier clinical intervention for infants at risk of developmental delays
• Demonstrates how AI can extract meaningful patterns from complex neurological data

Medical Impact: This research opens new possibilities for personalized pediatric care by identifying developmental concerns before traditional behavioral assessments can detect them, potentially improving long-term outcomes through early targeted interventions.

Swin fMRI Transformer Predicts Early Neurodevelopmental Outcomes from Neonatal fMRI