Neural development of high-level visual processing

How structural and functional features of the brain develop to support high-level visual processing is an important question in the extended neuroscience field. In collaboration with Kalanit Grill-Spector and the VPNL, we have recently shown that the topological layout of some large-scale maps is similar in children, adolescents, and adults. Specifically, functional representations that are biased toward the center of the visual field are cortically positioned on the lateral side of the mid-fusiform sulcus (MFS), while functional representations that are biased toward the peripheral portions of the visual field are positioned medial to the MFS. Thus, the MFS is a useful functional landmark across development.

Inflated cortical surface of a 7 year old child zoomed on ventral temporal cortex (left). Warm colors indicate those pieces of cortex that prefer the central (foveal) portion of visual space while the cool colors indicate those pieces of cortex that prefer the peripheral portion of visual space. The mid-fusiform sulcus (MFS) identifies the functional transition in this map in children, adolescents, and adults just 5mm from the fundus, or bottom, of the MFS (right). This large-scale structural-functional coupling across age groups may be a foundational building block for high-level visual representations. Image adapted from Weiner et al., 2013.

Interestingly, in both children and adults, face-selective regions are embedded within this large-scale foveal representation while place-selective regions are embedded within this large-scale peripheral representation. Most recently, we showed that microstructural changes occur within face-selective regions that are linked to the recognition memory of faces. Rather than pruning, we showed that it was more likely proliferation that contributed to these changes.

T1 relaxation in face- but not place-selective regions, negatively correlates with recognition memory. Left: Correlation (line and 95% confidence interval from bootstrapping) between mean recognition memory of faces vs. mean T1 in right pFus-faces. Right: Recognition memory for places vs. mean T1 in right CoS-places. In all plots, each point is a subject. Adapted from Gomez et al., 2017.

The CNL and VPNL will continue to collaborate on projects regarding the neural development of high-level visual processing. We are presently examining (1) the development of sulci and the positioning of cortical regions, (2) the microstructural development of vertical connections within face and reading networks, and (3) differences in cortical thinning and how it relates to myelination and development of functional regions within gyri vs. sulci.

Click to head back to Research page