Fig. 1
From: ID-Seg: an infant deep learning-based segmentation framework to improve limbic structure estimates
Overview of this study. a Using 3 independent infant MRI datasets through a transfer-learning approach, we trained, fine-tuned, and cross-validated a deep-learning segmentation framework (ID-Seg) for hippocampus and amygdala, both with internal and external datasets; b we further explored the prospective associations between morphometric measures (left and right hippocampus and amygdala) in infants and behavior problems at age 2. *Cyan color represents segmented hippocampus, and red represents segmented amygdala. LOOCV leave-one-out cross-validation