关键词: 自然阅读, 注意, 儿童, 脑电/脑磁图, 光生物调节技术

Abstract: This project proposes a comprehensive developmental cognitive neuroscience framework to investigate how school-age children encode parafoveal information during natural Chinese reading and how such attentional mechanisms shape reading acquisition. Although parafoveal preview and pre-saccadic attention have been widely studied in skilled adult readers, their neural underpinnings in children—especially within non-alphabetic scripts—remain largely unknown. To address this critical gap, the project advances a set of three integrative studies that combine naturalistic eye-tracking paradigms, wearable optically pumped magnetoencephalography (OPM-MEG), and transcranial photobiomodulation (tPBM) to elucidate the dynamic, hierarchical, and developmentally sensitive mechanisms of parafoveal attention. Together, these studies aim to build a multilevel theoretical model of how children’s brains coordinate foveal semantic processing with parafoveal orthographic encoding and how such coordination relates to reading fluency, attentional development, and potential intervention targets.
Study 1 focuses on identifying the core neural computations underlying children’s parafoveal attention during natural sentence reading. The central question is whether parafoveal information is encoded at orthographic or semantic levels before the eyes fixate on an upcoming word, and how these pre-saccadic representations relate to children’s eye-movement patterns, including fixation durations and saccadic step-length distributions. Using a gaze-contingent natural reading paradigm, OPM-MEG will record fixation-locked neural dynamics with millisecond precision while children read silently. By aligning MEG data with eye-tracked fixation events, the study will examine oscillatory phase patterns that reflect the depth of parafoveal processing. Multivariate representational similarity analysis (RSA) will be used to determine whether neural activity prior to fixation carries information about upcoming words’ orthographic features or semantic properties. The expected outcome is that younger children will show primarily orthographic-level parafoveal encoding, whereas older children may begin to exhibit emerging semantic-level preview effects. Furthermore, children whose eye movements more closely approximate a Lévy-flight distribution—a signature of efficient exploration—are expected to show deeper pre-saccadic processing and more mature neural oscillatory patterns. These findings will provide foundational evidence of the computational hierarchy through which children allocate attention across foveal and parafoveal regions.
Study 2 examines how the neural and behavioral markers identified in Study 1 vary across developmental stages and whether children exhibit age-dependent changes consistent with the proposed multiplexing phase-coding model. This model posits that different linguistic levels (orthographic vs. semantic) can be simultaneously represented in distinct oscillatory phases of the same fixation, enabling a “pipeline” architecture for efficient reading. The study compares younger (7-9 years) and older (10-12 years) children and employs rapid invisible frequency tagging (RIFT) to embed high-frequency flicker signals into parafoveal target words. By computing coherence between the flicker frequency and MEG signals during pre-saccadic windows, the study will quantify the depth and spatial extent of parafoveal attention. Larger coherence differences between valid and invalid previews will indicate deeper encoding. It is expected that with increasing age, children will display (a) larger parafoveal attention breadth, (b) more robust orthographic preview benefits, and (c) emergent semantic-level sensitivity. At the neural level, older children are predicted to show stronger alpha-phase lateralization and more reliable phase-amplitude coupling within frontoparietal networks, indicating maturation of attentional timing mechanisms. Together, these findings will reveal age-related shifts in the hierarchical organization of children’s reading-related attention networks and identify neural signatures marking the transition from “learning to read” to “reading to learn.”
Study 3 aims to establish causal evidence for the involvement of the frontal eye field (FEF)—a key node of the frontoparietal attention network—in modulating parafoveal attention during natural reading. Based on the neural markers revealed in Studies 1 and 2, this study applies tPBM targeting the FEF to examine whether enhancing local cortical excitability can increase the depth of parafoveal processing and improve reading fluency. In a randomized, double-blind, sham-controlled design, children will complete natural reading tasks after real or sham stimulation while undergoing eye-tracking and OPM-MEG recordings. The expected outcomes are that real stimulation will (a) strengthen alpha-phase consistency associated with pre-saccadic attention, (b) enhance the coherence between frequency-tagged parafoveal signals and neural activity, and (c) shorten fixation durations or promote more adult-like saccadic patterns. Such results would support the causal role of the FEF in orchestrating the temporal coordination between foveal and parafoveal processing. Additionally, younger children may show greater intervention benefits due to higher neural plasticity, providing insights into sensitive developmental periods for attention-based reading interventions.
The theoretical significance of this project lies in proposing and empirically grounding a developmental extension of the multiplexing phase-coding model for reading. By integrating naturalistic tasks, oscillatory dynamics, and developmental comparisons, the project advances a hierarchical-dynamic account of how children’s brains coordinate attention across multiple visual and linguistic layers. This work extends adult reading theories into a developmental framework and clarifies how general attentional maturation interacts with domain-specific literacy acquisition. Practically, the project builds a foundation for brain development-based reading education and opens avenues for safe, non-invasive, and individualized interventions for reading difficulties. By identifying neural markers that jointly reflect moment-to-moment attention allocation and reading ability, the project provides objective indicators for early screening of at-risk children. The incorporation of OPM-MEG and tPBM further lays the groundwork for closed-loop neural modulation systems tailored to children’s neural states. Overall, this research proposal integrates multi-modal neuroscience and developmental psychology to construct a comprehensive scientific basis for enhancing children’s reading outcomes and advancing national goals in brain and cognitive development.

Key words: natural reading, attention, children, EEG/MEG, Transcranial Photobiomodulation