The Neural Correlates Underlying Abacus Mental Calculation And The Brain Plasticity During Training Of Abacus Mental Calculation In Children

Abacus mental calculation (AMC),'Zhuxinsuan'in Chinese, is a sort of mental calculation based on the principles of manipulating an abacus. After training, child abacus learner has extraordinary capacity of mental calculation and digit memory. It is suggested that the training of AMC could develop children's ability of mathematics; might improve their cognitive function, and also may be helpful for young children who have problems in computation and concept in mathematics. Therefore, in China, learning of AMC captures not only the interests of children but also the attention of parents, educates and researchers. However, till now the neural correlates and cognitive mechanisms of AMC remain unknown.Functional magnetic resonance imaging (fMRI) is the most widely functional neuroimaging techniques to measure brain function with a temporal resolution of seconds and a good spatial resolution, and it has opened a new window into investigation of neural correlates underlying AMC and the brain plasticity during training of AMC in children.In this thesis, we try to investigate the following questions by fMRI experiments: (1) the neural correlates underlying AMC and abacus imagery; (2) the changes of brain activity pattern in response to long-term abacus training; (3) the modulation of the brain network and default mode network during AMC.The thesis has four main parts. Chapter one:Introduction, Chapter two:Neural correlates underlying abacus mental calculation, Chapter three:Plasticity of brain cortex and brain networks during AMC in children and last part:Conclusion and Future works. The introduction is concerned with the basics of AMC, common mental calculation, and brain plasticity, and with the review of previous studies about abacus learning and mental calculation. The introduction is also concerned with the basic knowledge of fMRI and functional connectivity. The second chapter identifies neural correlates underlying serial simple and complex abacus mental calculations, and explores the abacus imagery, by using the general linear model (GLM) and coherence functional connectivity analysis. The third chapter focuses on the relationship between abacus training and brain plasticity. And we investigate the issue by examining (a) relationship between plasticity and brain activity patterns during AMC, (b) the relationship between plasticity and neural network of AMC by using effective connectivity analysis based on structure equation modeling (SEM), and (c) the relationship between plasticity and intrinsic default mode network of abacus learners by functional connectivity analysis.In our first experiments a block-design fMRI experiment was carried out. We observed that (1) there are different neural correlates and neural networks underlying the AMC and traditional mental calculation, which may contribute to different strategies developed by the long-term training of abacus and AMC; and no classical language areas of the left hemisphere are involved in the processing of AMC; (2) child abacus experts primarily share the underlying dissociated neural correlates--the fronto-temporal circuit and the fronto-parietal circuit--respectively in simple and complex serial abacus; while in child controls almost the same brain network sustain the simple and complex tasks; (3) AMC may put emphasis on visuospatial-specific representations, and rely on the posterior temporal and posterior superior parietal areas in which abacus imagery can be generated by transforming the serial numbers into a super-modal form of abacus beads. These findings give us some lights that probably the training of AMC may be helpful for some young children who have difficulty in mathematics because of lesion in language areas.In our second experiment, child abacus experts performed AMC by both visual and auditory stimuli. We observed that the brain activity mapping was closely similar with two different stimuli, which confirmed that the neural correlate of AMC depended on the nature of the task rather than on the nature of the stimulus, and further supported that the activity in posterior temporal and posterior superior parietal areas did not reflect the visual input itself.In our third experiment, we try to explore the relationships between brain plasticity and the training of AMC in children by taking together several different methods including general linear model (GLM), ROI-based time course analysis, structural equation modeling (SEM) and correlation functional connectivity. We found that (1) the brain activity areas and neural response reduced in abacus children; (2) the brain network and effective connectivity paths of mental calculation changed after the training of AMC; (3) and the default mode network was modulated by training of AMC in children.It is the first time, to our knowledge, that the neural correlates underlying AMC and the brain plasticity during training of AMC in children have been reported. We hope the findings of the present studies would enrich the theories of number cognitive processing, and provide a solid foundation for and a direction to future researches.