应用数学和力学2025，Vol.46Issue(7)：836-854,19.DOI:10.21656/1000-0887.450234
头朝向细胞三维空间活动模式的神经动力学建模
Neurodynamical Modeling of 3D Spatial Activity Patterns of Head-Direction Cells
摘要
Abstract
Head-direction cells are present in several brain regions,including the mammalian medial entorhinal cortex,and respond selectively to specific head directions,and constitute a compass system in the brain.This system can update internal direction representations in a self-organized manner,and can receive inputs from the external environment to calibrate direction encoding.Currently,many computational models for head-direction cells only consider head directions encoding in the horizontal plane.Whereas experiments show that neurons encoding both horizontal azimuth and vertical pitch angles exist in the brains of mammals,there is a lack of computational modeling of their neural mechanisms.A continuous attractor network model was constructed to encode 3D direction features such as azimuth and pitch angles at the same time,realizing both the specific di-rection preference encoding of 3D-head-direction cells at the single-neuron level and the accurate tracking of head direction changes in 3D space at the population level.The torus topology used in the model,compared with the spherical topology,can more reasonably explain the neuronal tuning data for azimuth recorded by bats.The proposed neurodynamic model reproduces the phenomena encoded by electrophysiological experiments re-corded in 3D head directions and gives a mechanistic explanation of the dynamical angles of the activity patterns of head-direction cells in the 3D space.关键词
空间导航/神经动力学/头朝向细胞/连续吸引子网络/三维空间Key words
spatial navigation/neurodynamics/head-direction cell/continuous attractor network/3D space分类
数理科学引用本文复制引用
许霜,王毅泓,徐旭颖,潘晓川,王如彬..头朝向细胞三维空间活动模式的神经动力学建模[J].应用数学和力学,2025,46(7):836-854,19.基金项目
国家自然科学基金(12172132 （）
12272136:12472054) （）
