论文部分内容阅读
脑是支配人和高级动物活动的司令部和信息中心,脑电波是脑科学基础理论研究的重要指标,也是人类思维活动的体现.大量研究都表明脑电波信号与认知等功能活动有关.而脑电波信号是无数神经放电的混合,不同行为产生的电信号的强弱差别也很大,所以通常的盲源分离技术误差较大,很难将脑电波较好的分离出来并与动作相对应.通过建立数学模型,分析了脑电波与行为之间的联系.首先预测并验证了呼吸动作与脑电波之间关系模型,使用DFA模型经过最小二乘法拟合进一步确认了LFP变化的周期性.然后通过计算脑电波曲线的自相关与互相关,得到了脑电波的周期性以及呼吸相关脑电波成分周期性的相关关系.之后建立了独立成分分析(ICA)的模型使用InfoMax的方法对脑电波进行了成分分离,并通过对照实验比较、优化了分离结果.寻求最优的脑电波的分离模型.并在此基础上研究了小鼠的视觉感受区电位信号与视觉刺激间的关联.
The brain is the command and information center that governs the activities of humans and advanced animals. Brainwaves are an important indicator of the basic theory of brain science as well as an expression of human thinking activities. Numerous studies have shown that brain signals are related to functional activities such as cognition The signal of radio wave is a mixture of numerous nerve discharges, and the difference of the electric signal generated by different behaviors is also very large. Therefore, the error of the conventional blind source separation technique is relatively large, and it is difficult to separate the brain wave better and correspond to the movement. The relationship between brain wave and behavior was analyzed by establishing a mathematical model.Firstly, the model of the relationship between respiratory motion and brainwave was predicted and validated, and the periodicity of LFP change was further confirmed by the least square fitting using DFA model. By calculating the autocorrelation and cross-correlation of EEG curves, the periodicity of EEG wave and the periodic correlation of respiratory-related EEG components were obtained.Afterwards, a model of independent component analysis (ICA) was established by using InfoMax The components were separated and the results of the separation were optimized by comparing with the control experiment.Secondly, the best separation model of brain waves was sought. We studied the correlation between the visual perception region of the mouse with the potential signal on the basis of visual stimuli.