Functional brain networks underlying the interaction between central and peripheral processes involved in Chinese handwriting in children and adults

Abstract

The neural mechanisms that support handwriting, an important mode of human communication, are thought to be controlled by a central process (responsible for spelling) and a peripheral process (responsible for motor output). However, the relationship between central and peripheral processes has been debated. Using functional magnetic resonance imaging, this study examined the neural mechanisms underlying this relationship in Chinese handwriting in 36 children (mean age = 10.40 years) and 56 adults (mean age = 22.36 years) by manipulating character frequency (a central variable). Brain network analysis showed that character frequency reconfigured functional brain networks known to underlie motor processes, including the somatomotor and cerebellar network, in both children and adults, indicating that central processing cascades into peripheral processing. Furthermore, the network analysis characterized the interaction profiles between motor networks and linguistic-cognitive networks, fully mapping the neural architecture that supports the interaction of central and peripheral processes involved in handwriting. Taken together, these results reveal the neural interface underlying the interaction between central and peripheral processes involved in handwriting in a logographic writing system, advancing our understanding of the neural basis of handwriting.

Keywords: brain network; central processing; handwriting; interaction; peripheral processing.

FIGURE 1

Descriptive statistics and analysis of in‐scanner behavioral performance. The writing duration (a) and latency (b) of children and adults while copying Chinese characters. HFCs, high‐frequency characters; LFCs, low‐frequency characters. n.s., not significant. *p < .05, **p < .01

FIGURE 2

Functional brain networks that differ between the conditions of copying HFCs and LFCs. The brain plots show the functional brain network with greater connectivity in the HFC condition than the LFC condition in children (a) and adults (b). The colors of the nodes in the brain plots indicate the network to which they belong. The large nodes represent hubs, whose sizes are proportional to the node strengths. The matrix plots represent connectivity strength between pairs of the 12 brain networks in children (c) and adults (d). The color of each element in the matrices represents the sum of the weight of all the edges for the connected networks. The patterns of intranetwork connectivity within the SMN and internetwork connectivity between the SMN and cerebellar network in children (e) and adults (f) are illustrated. AN, auditory network; Cereb, cerebellar network; CON, cingulo‐opercular network; DAN, dorsal attention network; DMN, default mode network; FPN, frontal–parietal network; HFCs, high‐frequency characters; L, left; LFCs, low‐frequency characters; R, right; SAN, salience network; SCN, subcortical network; SMN, somatomotor network; Unc, uncertain; VAN, ventral attention network; VN, visual network; Z, Fisher's z scores