先天性白内障手术后视功能重塑的研究进展
Research progress on visual function remodeling after congenital cataract surgery
摘要
先天性白内障是我国儿童致盲的主要病因之一。早期手术虽能解除解剖结构障碍,但手术后视功能重塑疗效欠佳,核心原因是不同个体的视觉通路中,Hebbian可塑性机制与内稳态可塑性机制的功能状态及二者协同代偿能力存在差异。Hebbian可塑性作为正反馈机制,通过长时程增强与抑制实现突触精细化修剪;内稳态可塑性作为负反馈机制,维持神经网络的整体稳态。二者在前、后视觉通路中通过不同的分子机制发挥交互作用,确立了"低阶补偿+高阶训练"联合治疗策略的神经生物学基础。传统的康复策略包括屈光矫正、基于内稳态机制的遮盖压抑疗法和基于Hebbian机制的弱视治疗仪,因缺乏任务导向性,未能兼顾两种机制的独立作用与交互效应,临床应用疗效受限。而视感知觉学习可精准整合"快速Hebbian扰动-慢速内稳态补偿"的动态交互模式,利用双眼分视、单眼刺激及跨模态整合等范式,结合虚拟现实、增强现实等数字环境,显著提升患儿训练依从性与视功能重塑效果。现有研究仍存在机制探索多依赖动物模型、疗效评价指标未统一及数字智能环境长期安全性不明确等局限。未来需开展多中心纵向研究,推进治疗方案标准化,构建精准评估与干预体系,以实现个体化治疗并改善患儿视功能预后。
Congenital cataract is a leading cause of childhood blindness in China. While early surgical intervention eliminates the anatomical obstruction, the efficacy of postoperative visual function remodeling remains suboptimal. This is primarily attributed to individual heterogeneity in the functional status and synergistic compensatory capacity of Hebbian and homeostatic plasticity within the visual pathway. Hebbian plasticity acts as a positive feedback mechanism to achieve refined synaptic pruning through long-term potentiation and depression, while homeostatic plasticity serves as a negative feedback mechanism to maintain the overall homeostasis of neural networks. The two exert interactive effects via distinct molecular mechanisms in the anterior and posterior visual pathways, which establishes the neurobiological basis for the combined treatment strategy of "low-order compensation+high-order training". Traditional rehabilitation strategies, including refractive correction, occlusion and penalization therapy based on homeostatic plasticity, and amblyopia therapeutic instruments based on Hebbian plasticity, have limited clinical efficacy due to the lack of task-oriented design and the failure to address both the independent roles and interactive effects of the two mechanisms. In contrast, visual perceptual learning can precisely integrate the dynamic interaction mode of"rapid Hebbian perturbation-slow homeostatic compensation". By adopting paradigms such as dichoptic presentation, monocular stimulation and cross-modal integration, combined with digital environments including virtual reality and augmented reality, visual perceptual learning significantly improves the training adherence and visual function remodeling outcomes of pediatric patients. Current studies still have several limitations: mechanistic exploration mostly relies on animal models, unified efficacy evaluation indicators have not been established, and the long-term safety of digital intelligent environments remains unclear. In the future, multicenter longitudinal studies should be conducted to advance the standardization of treatment regimens, construct a precise evaluation and intervention system, so as to realize individualized treatment and improve the visual function prognosis of affected children.
年,卷(期):
2026, 42(10)
在线出版日期:
2026年10月25日