| How the brain generates advanced cognitive functions has been a subject of ongoing exploration.Studies have demonstrated that neural information transmission between cortical regions and neuronal populations is crucial for advanced cognitive functions.However,the complex and nonlinear nature of the brain’s nervous system hinders a complete understanding of the mechanism of neural information transmission.Therefore,exploring how neural information is transmitted within the brain is a key scientific question in the field of neuroscience.Yet,the experimental study of neural information transmission faces considerable challenges due to technological limitations.Researchers have found that by combining experimental data with neuronal network modeling,it is possible to overcome these experimental limitations.It enables flexible and systematic investigation of neural information transmission,leading to valuable research progress.The brain’s nervous system exhibits a modular network structure.In previous studies on neural information transmission mechanisms using dynamic modeling,researchers typically employed feedforward neuronal network models to capture the network topology.They relied on experimental data to examine neural coding transmission.However,with the development of technology,experimental personnel are gradually able to explore the brain’s structure and measure neural activity more flexibly at different scales.Therefore,further investigation into the mechanism of neural information transmission becomes necessary,building upon previous theoretical research and incorporating new experimental evidence.This thesis aims to integrate recently discovered experimental evidence and investigate information transmission in neuronal networks.It adopts a computational modeling and dynamic perspective to analyze the underlying neural mechanisms involved in information transmission.Furthermore,this study specifically focuses on exploring the potential mechanisms of neural information transmission in feedforward neuronal networks.The main results and conclusions of this thesis include the following three parts:1)A new mechanism for encoding and transmission of neural information in a mixed excitatory-inhibitory synapse neuronal network structure.With the advancement of technology,researchers have observed that inhibitory and excitatory neurotransmitters are coreleased from the same synapse terminal in some brain regions.Inspired by this,This thesis o proposes a novel neuronal network with mixed excitatory and inhibitory synapses and builds upon it a feedforward neuronal network to investigate information encoding and transmission within it.The results show that under identical conditions of parameters,the proposed network has higher coding efficiency than traditional networks.This thesis then studies the transmission of population firing rate and pulse packets and discovers that the proposed network can respectively support the stable transmission of population firing and pulse packets under different network states.The research results elucidate the encoding transmission mechanism of mixed excitatory inhibition in synaptic neural networks,which helps to understand the impact of mixed synaptic neurotransmitters on information transmission.2)Multi-signal transmission mechanism in neuronal networks.During cognitive tasks,neuronal networks in the brain need to process and transmit different signals simultaneously,but the underlying neural mechanism is still not completely clear.This thesis proposes frequency division multiplexing(FDM)as the basic mechanism to support multiple signal transmission simultaneously.A two-layer feedforward neuronal network(sender-receiver)is constructed to simulate the upstream and downstream brain regions,studying how multiple signals can be simultaneously transmitted in the neuron system.Each independent network possesses two intrinsic working rhythms of slow and fast gamma.The results show that the neuronal network can transmit low-and high-frequency signals simultaneously through the FDM communication mechanism.In addition,precise regulation of network state can enable selective transmission of different frequency signals.The research findings may provide new insights into the potential mechanisms for the simultaneous transmission of complex signals between different cortical regions,helping to further understand the cross-brain communication mechanisms of complex signals in cognitive processes.3)Subspace communication mechanism in feedforward neuronal networks.Recent experimental evidence shows that task-related activity is limited to a low-dimensional space(manifold),and different brain regions can transmit neural activity through such subspace.However,it is currently not clear under what conditions this subspace communication is feasible in cortical spiking neuronal networks.This thesis constructs a feedforward neuronal network to achieve subspace communication and studies its reliability.The results show that the proposed network structure can transmit the activity in the potent subspace to the target network,while the activity in the null subspace is blocked in the target.Additionally,it is found that the reliability of this subspace communication is negatively related to the input frequency,spatial correlation,and sequence correlation.The research results indicate that subspaces may be an effective mechanism for supporting flexible and selective gating of neural information.It helps to further understand the mechanism of interareal communication and provides a theoretical basis for future research on specific neural circuits and related cognitive functional principles.In summary,referring to the recent experimental evidence,this thesis has carried out indepth research on the potential neural mechanisms of information coding and transmission in mixed synapse neuronal networks,multi-signal simultaneous transmission,and information selective gating based on subspace communication.The research findings reveal the potential neural mechanisms of communication between neuronal networks in the brain.It also helps to further understand how information is encoded and transmitted in neuronal populations and networks.Additionally,the research findings may also provide certain models and theoretical support for understanding the neural computational principles of cognitive processes,as well as the mechanisms and treatment of neural diseases. |
