| Summary: | 博士 === 國立臺灣大學 === 心理學研究所 === 102 === Purpose: functional magnetic resonance imaging (fMRI), dynamic causal modelling (DCM) hierarchical model comparisons, and diffusion spectrum imaging (DSI) were used to investigate the effective connectivity and structural connectivity during semantic judgments to visual Chinese characters. All participants were asked to indicate if character pairs were related in meaning. Methods: In Experiments 1 and 2, the use of DCM was to examine the directional influences among brain regions, and to seek for the optimal model in the semantic network by hierarchical model comparisons. In adult participants, the experimental stimuli were character pairs that included semantically-related and semantically-unrelated pairs. In Experiment 3, Bayesian Model Selection (BMS) and Bayesian Model Averaging (BMA) were used to seek for the optimal model. In 17 children (10- to 13-year-olds) and 17 adults, the experimental stimuli included high association, low association, and unrelated pairs. In Experiment 4, fMRI, DSI and DCM were used to study the developmental changes of neural correlates, structural connectivity, and effective connectivity during semantic judgment. Results: In conventional fMRI analysis, common activation was found in left ventral inferior frontal gyrus (IFG, BA 45, 47), left posterior middle temporal gyrus (MTG, BA 21) in Experiments 1 and 2. In Experiment 3, both groups showed greater activation in left ventral IFG, left MTG, and left fusiform gyrus (FG, BA37). In Experiment 4, compared to children, adults showed greater activation in the left ventral IFG and MTG, and adults had significantly greater structural connectivity in the left ventral pathway (inferior frontal occipital fasciculus, IFOF) than children. As for effective connectivity from DCM analyses, significant modulatory effects were found from left ventral IFG to left MTG, from left MTG to left ventral IFG, and from left FG to left MTG in Experiment 1. In Experiment 2, the results of the optimal model showed significant modulatory effects from ventral IFG to MTG and from FG to MTG. In Experiment 3, a significant bottom-up effect was found from FG to MTG. Moreover, a bottom-up effect from FG to ventral IFG was stronger than a top-down effect from ventral IFG to FG in adults, and that the bottom-up effect from FG to ventral IFG was stronger in adults compared to children. In Experiment 4, adults showed significantly stronger bottom-up connection from FG to ventral IFG than children in the related condition. Conclusion: First, the modulatory effects from left ventral IFG to left MTG suggested top-down influences of the frontal cortex on retrieval of semantic representations. Second, the modulatory effects from left MTG to left ventral IFG suggested the role of MTG on providing relevant associations in verbal semantic memory for ventral IFG to perform retrieval. Third, the modulatory effect from left FG to left MTG suggested bottom-up orthographic influences on semantic representations. The most important finding was the bottom-up effect from FG to ventral IFG in adults, implying developmental changes in semantic processing. In conclusion, our findings suggest that age-related maturation in brain activation (ventral IFG and MTG) and structural connectivity (IFOF) might be associated with the bottom-up influence of orthographic representations on retrieving semantic representations for processing Chinese characters.
|
|---|
