Information Encoding Mechanism Of Place Cell Functional Network In Goal-directed Behavior Of Pigeons

Spatial cognition and motion navigation are the basic functions of brain and are essential for the survival of animals.Most of the biological spatial navigation in nature is goal-directed,such as foraging,homing,courtship,migration.Research on the neural mechanism of goal-directed behavior has become one of the interdisciplinary frontier of neuroscience and control science.In recent years,a series of neurophysiological experiments show that the hippocampus(Hp)area plays a key role in the goal-directed behavior.Place cells in the Hp are considered to be the basic unit of spatial information processing,whose discharge is strongly related to a animal's location.And specific firing sequence of different place cells reflects the path.However,it is unclear how to achieve function interconnection among place cells,then coordinate the encoding of key navigation information,such as goal and routing.Thus,based on the in-depth analysis of functional connection among place cells,clarifying information encoding mechanism of place cell network(PCN)for the goal-directed behavior is the key of the biological spatial navigation neural mechanism research.At the same time,the study also has essential applied value for bionic navigation and robotics.To reveal the encoding mechanisms of functional PCN on the goal-directed behavior,this study uses pigeon with excellent spatial cognition and navigation ability as a model animal.Firstly,we record the in vivo neural signals from the Hp using the microelectrode array,compare the spatial response characteristics of place cells in the tasks with and without goal,and constructe the spatial cognitive map.Secondly,using the complex network theory and method,the functional PCN is structured and its topological characteristics is analyzed in goal-directed spatial cognitive task.Then the encoding mechanisms of functional PCN on both the goal-information and the routing-information are studied,and the prediction performance of functional PCN for the routing information is analyzed in detail.Lastly,the state-space encoding model of functional PCN is established and its validity is verified by the motion trajectory neural coding and decoding experiments.The main contents and innovation points of this paper are presented as follows:1)Developing a pigeon brain stereotactic device and a microelectrode implantation depth dynamic adjustable device,which solve the precise positioning problem of the microelectrode array and provide a reliable guarantee for reliable detection of the in vivo neural signal and long-term stable recording.For the unique anatomical structure of bird's head,a new stereotaxic apparatus for pigeon's brain is developed by the four-points locating method to achieve the precise positioning of microelectrode array implantation.For the in vivo microelectrode array implantation,a depth dynamic adjustable device is made using the modular design to improve the success rate of microelectrode implantation and extend the acquisition time of neural signal.2)Through the pigeon spatial cognitive experiment analysis,we find that the place cells are more likely to form a stable place field in the goal-directed spatial cognitive task.The place field characteristics of place cells are compared in the both tasks with and without goal from the spike characteristics,the spatial characteristics and the stability aspects,we find that the place fields are more reliable in the tasks with goal than that without goal.And different from the mammals,the place cells of pigeon generally have multiple place fields feature.It indicates that the activation of place fields is strongly dependent on the modulation of goal-directed information.3)The two functional PCNs are respectively structured by the spike firing rates and the LFP theta rhythm features.Through the analysis of functional PCN's topological characteristics,we find that the place fields are highly clustered in the goal-directed spatial cognitive task.And the clustering coefficient and global efficiency of network in the place field areas are significantly higher than both that in the non-place field areas,but the small world characteristic of network in the place field areas is significantly lower than that in the non-place field areas.Indicating that the functional connections among place cells become more closely for encoding of a specific position in the goal-directed behavior,which helps to increase the stability of goal-directed.4)Through the analysis of functional PCN,we find that the Hp and NCL neurons play a key role in the goal information and routing information encoding in goal-directed decision-making task,respectively.The functional network of Hp place cells mainly encodes the current position and goal position information,and is not sensitive to the routing information.And the functional network of NCL neurons mainly encodes the routing information.The results preliminarily reveal the role and division of the Hp and the NCL in the goal-directed behavior of avian.5)Based on the functional connections among place cells,the state-space encoding model of functional PCN is established by first order random walk model and quadratic exponential Poisson's equation.And the neural encoding and decoding experiments of motion trajectory are carried out.The prediction results of model are consistent with the measured results,which verifies the validity of model.Moreover,the particle filtering algorithm is introduced into the motion trajectory prediction in neural decoding experiment,which improves the accuracy of trajectory prediction.