基于生理连续性的WZ神经元模型改进及视网膜神经节细胞动力学仿真

Improvement of WZ Neuron Model Based on Physiological Continuity and Dynamic Simulation of Retinal Ganglion Cells

摘要: 神经元动作电位动态机制是神经电生理学与计算神经科学的核心研究内容，构建高生理保真度、低计算复杂度的量化模型是解析其内在规律的关键。针对传统WZ模型存在阈下兴奋过程缺失、生理适配性偏低、连续放电场景受限、优化后计算复杂度剧增等缺陷，本文以生物电生理连续性为导向对WZ模型进行系统性改进。将动作电位划分为阈电位触发、去极化上升、复极化下降、超极化恢复四个阶段，建立分段参数调控与跨阶段连续性约束，引入时间重置机制保证动作电位“全或无”特性；在此基础上开展连续时序刺激下神经元放电仿真，并构建4 × 4感光阵列与侧向抑制模型，实现视网膜神经节细胞拮抗感受野动力学模拟。结果表明，改进模型可完整复现动作电位全周期波形，严格遵循电生理规律，在连续刺激下放电节律稳定；仿真结果准确复现给光中心–撤光周边的经典感受野特性，弥散光照下响应微弱，与生理实验一致。本文模型在保留计算简洁性的同时显著提升生理真实性与场景适用性，为神经元电信号解析、视网膜视觉通路建模与神经动力学仿真提供有效工具。

Abstract: The dynamic mechanism of neuronal action potentials constitutes the core research content in neuroelectrophysiology and computational neuroscience. Constructing quantitative models with high physiological fidelity and low computational complexity is crucial for analyzing its intrinsic laws. Aiming at the defects of the traditional WZ model, such as the absence of subthreshold excitation processes, low physiological compatibility, restricted continuous firing scenarios, and a sharp increase in computational complexity after optimization, this paper systematically improves the WZ model guided by biological electrophysiological continuity. The action potential is divided into four stages: threshold potential triggering, depolarization rising, repolarization falling, and hyperpolarization recovery. Piecewise parameter regulation and cross-stage continuity constraints are established, and a time reset mechanism is introduced to ensure the “all-or-none” characteristic of action potentials. On this basis, neuronal firing simulations under continuous sequential stimulation are carried out, and a 4 × 4 photoreceptor array together with a lateral inhibition model is constructed to realize the dynamic simulation of the antagonistic receptive field of retinal ganglion cells. The results show that the improved model can fully reproduce the full-cycle waveform of action potentials, strictly follow electrophysiological laws, and maintain stable firing rhythms under continuous stimulation. The simulation results accurately replicate the classical receptive field characteristics of ON-center/OFF-surround, with weak responses under diffuse illumination, which is consistent with physiological experiments. While retaining computational simplicity, the model in this paper significantly improves physiological authenticity and scenario applicability, providing an effective tool for the analysis of neuronal electrical signals, modeling of retinal visual pathways, and neural dynamic simulation.