失匹配负波加工性质的研究

发布时间：2018-03-10 19:43

本文选题：失匹配负波　切入点：扩散模型　出处：《大连理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：失匹配负波(Mismatch Negativity, MMN)是一种声音变化检测机制,MMN可能与听力的预注意认知过程相关联。因此,有人认为它反映了在听觉皮层的“原始智能”,但其是否是脑的自动加工的指标,一直是MMN研究的争论问题,同时也是解释MMN形成机理的一个重要论据。尽管很多的实验范式都对MMN的自动与非自动加工性质提出了自己的解释,但目前鲜有实验范式能够很好的控制受试者的注意力,致使在非注意通路中会引入受试者细微的注意力资源的变化。为了解决这一问题,很多学者利用注意力缺失的患者进行实验,但因受试本身的大脑功能已经出现缺陷,而在临床上这一缺陷产生的原因尚不明确,无法确定这一缺陷会对哪些因素产生影响,因此在这种情况下得到的结果就不能反推出正常受试的结果。针对以上问题,本文拟对跨通路延迟反应实验范式进行改进,使实验范式能够更好的控制受试者的注意力资源,并在实验过程中采集受试者的行为数据,然后利用扩散模型对行为数据进行拟合,将拟合出的参数作为受试者注意力资源分配的指标。由于在ERP数据采集过程中信号的信噪比较低,引入了较多的干扰,对MMN的成分的提取产生了影响,因此本文采用传统叠加平均加、vavelet Indepandent Component Analysis算法来提取MMN成分。最后结合MMN的峰潜伏期与峰振幅,从而明确MMN与注意力间的关系,确定其加工的性质。经过实验与数据分析后发现,扩散模型所拟合出的行为数据参数可以解释实验设计中各条件对受试者注意力资源变化的潜在心理变化,定量地确定受试者注意力。此外MMN的峰潜伏期随实验条件变化显著,并且其与扩散模型的参数有显著的相关性：与边界间隔α的相关系数为r=-0.63,与非决策时间Ter的相关系数为r=0.63,与漂移率v的相关系数为r=0.58。结果证明扩散模型的参数可以作为MMN与受试者注意力资源分配的指标,且MMN的峰潜伏期随受试者注意力呈正相关性,因此我们认为MMN具有半自动加工的性质。本文利用受试者生理数据与行为数据的对MMN的加工性质进行研究的一个创新,这也为今后的其它ERP成分加工性质的研究提供了一条新的有益途径。
[Abstract]:Mismatched negative wave mismatch Negativity (MMN) is a sound change detection mechanism that may be associated with the prenotice cognitive process of hearing. Therefore, it is thought to reflect the "primordial intelligence" in the auditory cortex, but whether it is an indicator of automatic processing of the brain. It has been a controversial issue in MMN research and an important argument to explain the formation mechanism of MMN. Although many experimental paradigms have provided their own explanations for the automatic and non-automatic processing properties of MMN, there are few experimental paradigms that can well control the subjects' attention. In order to solve this problem, many researchers used attention-deficit patients to conduct experiments, but because the brain function of the subjects themselves has been defective, However, the causes of this defect in clinic are not clear, and it is impossible to determine which factors will be affected by the defect, so the results obtained in this case can not reverse the results of normal subjects. In view of the above problems, this paper intends to improve the experimental paradigm of cross-pathway delay response, so that the experimental paradigm can better control the attention resources of the subjects, and collect the behavior data of the subjects during the experiment. Then the diffusion model is used to fit the behavioral data, and the fitted parameters are used as indicators for the allocation of attention resources. Because of the low signal-to-noise ratio (SNR) of the signal in the process of ERP data acquisition, more interference is introduced. It has an effect on the extraction of MMN components. In this paper, the traditional superposition average plus Indepandent Component Analysis algorithm is used to extract MMN components. Finally, combining the peak latency and amplitude of MMN, the relationship between MMN and attention is clarified. Determine the processing properties. After experiment and data analysis, it is found that the behavioral data parameters fitted by diffusion model can explain the potential psychological changes of each condition in the experimental design to the change of the subjects' attention resources. In addition, the peak latency of MMN varied significantly with the experimental conditions. And it has significant correlation with the parameters of diffusion model: the correlation coefficient with boundary interval 伪 is r-0.63, the correlation coefficient with non-decision time Ter is 0.63, and the correlation coefficient with drift rate v is rn0.58. the results show that the parameters of diffusion model can be obtained. As an indicator of the allocation of attention resources between MMN and subjects, The peak latency of MMN is positively correlated with the subjects' attention, so we think that MMN has the property of semi-automatic processing. This paper makes use of subjects' physiological data and behavioral data to study the processing properties of MMN. It also provides a new and beneficial way to study the processing properties of other ERP components in the future.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.7;R318

【相似文献】