Separating group- and individual-level brain signatures in the newborn functional connectome: A deep learning approach

Jung-Hoon Kim, Josepheen De Asis-Cruz, Catherine Limperopoulos

Abstract: Highlights•The VAE shows significant subject identification accuracy with whole-brain FC profiles, surpassing the identification accuracy obtained with linear countermodels.•The VAE is able to reliably separate the neonatal FC profile into connections that track age and connections that reflect individual brain signatures.•Cortical functional network, derived by the VAE, varied in their representation of brain maturation and individual signatures; notably, certain CFNs that failed to capture neurodevelopmental traits, in fact, exhibited individual signatures.