Functional Analysis Of Neuron Function And Prediction Of Cognitive Behavior

Brain research is one of the most important subjects in21century. The brain information processing is a comprehensive interdisciplinary, which combine with the neurobiology, computer science, and mathematics. The traditional analysis methods cannot be applied to the analysis requirements of multi-electrode array recordings. Complex network theory, which combines with graph theory, has become the important part in neuroscience research. The existed researches have focused on the small-world properties of brain functional networks and mainly used for macro-level (magnetic resonance imaging, EEG) analysis. These networks cannot reflect the activity patterns of neurons. The researches from the micro-level can be more clearly used to describe the neuronal connectivity.This PhD thesis regard the neuronal functional networks as research object via recording the population neurons in rat using multi-electrode recordings in vivo. Combining with the computational model analysis, we study the characteristics of the neuronal functional networks, the division of the community structure, evaluation of the community partitioning, and using the functional networks to predict the behavioral choices of the rats. This research, not only provides the neural mechanisms to analyze the brain cognitive function, and also provides a new direction for the diagnosis and treatment of brain diseases.The main contents and results are organized as follows:Neuroscientists still lack of understanding of how the evolution of the brain functional networks when the animal performed cognitive tasks. This project constructs a computational model of hippocampal neuronal circuits according to the anatomical basis of the neurobiology of the hippocampus. We simulate the spiking of single neurons. Based on connection structure between neurons, we simulate the activities of neuronal populations and study the change of the connection structure. Since we cannot know the numbers and features of real neuronal functional networks in advance, the surrogate spike train data sets need to be generated and used for testing the validity of the proposed analysis method.In recent years, complex-network analysis methods, which are constructed via functional magnetic resonance imaging (fMRI), electroencephalogram, and magnetoencephalogram, have been widely used in analyzing the functional connectivity networks of the brain. These results were significant in analyzing the changes of brain connectivity structures and brain diseases. However, the topological characteristics of neuronal functional networks (NFNs) are barely understood because of the limited studies available on brain functional networks constructed based on the level of individual neurons. Therefore, this study recorded spike trains from brain cortical neurons of several behavioral rats in vivo by using multi-electrode recordings. The rats performed two different cognitive tasks. An NFN was constructed on each trial, obtaining a total of150NFNs in this study. The topological characteristics of NFNs were analyzed by using the three most important characteristics of complex networks, namely, small-world structure, scale-free network, and community structure. We found that the small-world properties generally exist in different NFNs. Compared with fMRI data networks, NFNs lack considerable scale-free behavior. A criterion was defined to determine the existence of community structure in NFNs, through which we found that community-structure characteristics, which are related to recorded spike train data sets, are more evident in the Y-maze task than in the DM-GM task. Results show that not all small-world functional networks have modularity and not all small-world functional networks could be partitioned into community structures, thus suggesting the effectiveness of the complex-network analysis methods and denoting that important complex network characteristics are present in NFNs obtained via multi-electrode recordings. Our results can also be used to analyze further the relationship between small-world characteristics and the cognitive behavioral responses of rats.Determining community structure in networks is fundamental to the analysis of the structural and functional properties of those networks, including social networks, computer networks, and biological networks. Modularity function Q, which was proposed by Newman and Girvan, is the most widely used criterion for evaluating the partition of a network into communities. However, modularity Q is subject to a serious resolution limit. In this paper, we propose a new function for evaluating the partition of a network into communities. This is called community coefficient C. Using community coefficient C, we can automatically identify the ideal number of communities in the network, without any prior knowledge. We demonstrate that community coefficient C is superior to the modularity Q and does not have a resolution limit. We also compared the two widely used community structure partitioning methods, the hierarchical partitioning algorithm and the normalized cuts (Ncut) spectral partitioning algorithm. We tested these methods on computer-generated networks and real-world networks whose community structures were already known. The Ncut algorithm and community coefficient C were found to produce better results than hierarchical algorithms. Unlike several other community detection methods, the proposed method effectively partitioned the networks into different community structures and indicated the correct number of communities.In this paper, we propose a new neuronal functional network community structure detection method. Firstly, we use the random walk distance to determine similarity between pair of neurons, and use the nearest neighbor method to sort the similarity matrix. Finally, we use the spectrum decomposition method to decompose the new similarity matrix. By analyzing the eigenvalue the eigenvectors, we automatically determine the number and structure of neuronal functional networks. We evaluated the performance of this method on surrogate data sets which know the structures in advance. We apply this method on the recorded spike trains in rat. The rat performed the Y-maze behavioral task. We find the community structures from neuronal networks. The traditional methods cannot find these structures.Analyzing the neuronal organizational structures and studying the changes in the behavior of the organism is key to understanding cognitive functions of the brain. Although some studies have indicated that spatiotemporal firing patterns of neuronal populations have a certain relationship with the behavioral responses, the issues of whether there are any relationships between the functional networks comprised of these cortical neurons and behavioral tasks and whether it is possible to take advantage of these networks to predict correct and incorrect outcomes of single trials of animals are still unresolved. This paper presents a new method of analyzing the structures of whole-recorded neuronal functional networks (WNFNs) and local neuronal circuit groups (LNCGs). The activity of these neurons was recorded in several rats using a multi-electrode recording system. The rats performed two different working memory behavioral tasks, the Y-maze task and the U-maze task. Using the results of the assessment of the WNFNs and LNCGs, this paper describes a realization process for predicting the behavioral outcomes of single trials. The methodology consists of four main parts:construction of WNFNs from recorded neuronal spike trains, partitioning the WNFNs into the optimal LNCGs using social community analysis, unsupervised clustering of all trials from each dataset into two different clusters, and predicting the behavioral outcomes of single trials. The results show that WNFNs and LNCGs have the direct effects on specific behavioral choices. The U-maze datasets show higher accuracy for unsupervised clustering results than those from the Y-maze task, and these datasets can be used to predict future behavioral responses effectively. The results of the present study suggest that methodology proposed in this paper is suitable for analysis of the characteristics of neuronal functional networks and the prediction of rat behavior. The WNFNs and LNCGs corresponding to two different behavioral choices exhibited obvious differences when the rats performed the U-maze task. These behavioral changes were found to be correlated with functional connections. These types of structures in cortical ensemble activity may be critical to information representation during the execution of behavior.