Back

Development of Sensorimotor-Visual Connectome Gradient at Birth Predicts Cognitive Outcomes at 2 Years of Age

Xia, Y.; Xu, Y.; Duan, D.; Xia, M.; Jeon, T.; Ouyang, M.; Chalak, L.; Huang, H.; He, Y.

2023-09-08 neuroscience
10.1101/2023.09.07.556778 bioRxiv
Show abstract

Functional connectome gradients represent fundamental brain organizing principles. Here, we investigate the maturation of the macroscopic connectome gradients in a task-free functional MRI dataset consisting of 39 preterm and term babies aged 31-42 postmenstrual weeks and its connection with postnatal cognitive growth. We show that the principal sensorimotor-to-visual gradient is present in baby connectomes around 35.6 postmenstrual weeks and continuously develops towards a more term-like pattern. The gradients global topographical structure undergoes progressive maturation with age, characterized by increases in explanation ratio, gradient range and gradient variation. The development of focal gradients is mostly localized in primary regions serving sensory, motor and visual functions. Machine learning approaches revealed that the connectome gradient at birth significantly predicts individual cognitive outcomes at 2-y follow-up. These findings highlight the early emerging rules of functional connectome gradients and have implications for understanding how the connectome gradients at birth underlies cognitive growth in later life. HighlightsO_LISensorimotor-to-visual gradient is present around 35.6 postmenstrual weeks C_LIO_LIGlobal gradient structure undergoes substantial changes during the third trimester C_LIO_LIRegional gradient develops mostly in primary sensory, motor, and visual regions C_LIO_LISensorimotor-to-visual gradient at birth predicts cognitive outcomes at age 2 C_LI

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.