脉络膜三维形态对儿童高度近视进展的影响研究

The incidence of high myopia among children and adolescence has been rising recently, and has become one of the most important causes of visual impairment. Choroidal atrophy plays a critical role during the development of high myopia and its related fundus changes. However, the changes of choroidal structure in the early stage of high myopia in children and adolescence remain unclear. In previous studies, we found that the choroid got thinner unevenly in different areas of the posterior pole, but it followed certain rules during the early stage of high myopia. This suggested that the spatial structure of choroid could change under similar rules and was associated with the development of high myopia. This study will be performed based on the cohort of high myopic children, which has been established in our former study. We will collect images of fundus photography and optical coherence tomography (OCT), with the application of denoising, automatic identification of choroidal structures and three-dimensional reconstruction technology, in order to realize the quantitative analysis of choroid. On this basis, we will construct mathematical models to find the major influence factors of choroidal three-dimension structure and to analyze its correlation with high myopia. This study will provide a new insight into the early fundus changes in high myopic eyes. At the same time, it will help to the prevention, intervention and control of high myopia.

高度近视在儿童青少年中的发病率逐年升高，是目前引起视力损害的重要原因。脉络膜萎缩变薄在高度近视及其相关眼底病变的发生发展中扮演着重要角色，但是在儿童高度近视病程早期，脉络膜内部空间结构的改变尚不清楚。前期研究中，我们发现高度近视儿童后极部眼底各区域脉络膜不均匀变薄，并且遵循一定的规律，这提示我们脉络膜内部三维空间结构的改变也可能存在类似的规律，并且与高度近视进展相关联。本课题拟基于课题组前期已建立的儿童高度近视登记研究队列，采集眼底彩照和光相干断层扫描图像，运用基于小波变换的混合滤波降噪、脉络膜结构自动化识别、三维重建等技术，实现脉络膜结构三维空间任意区域的定量分析；在此基础上，通过构建数学模型，寻找脉络膜三维结构的主要影响因素，并分析三维结构特征与高度近视发展的关联性。本课题实施将为探索高度近视早期眼底改变的发生机制提供新思路，同时为高度近视的进展控制和眼底病变的预防与干预提供重要参考。

高度近视在儿童青少年中的发病率逐年升高，是引起视力损害的重要原因。脉络膜结构的改变在高度近视及其相关眼底病变的发生发展中扮演着重要角色。在儿童高度近视病程早期，脉络膜内部空间结构已发生较为明显的改变，该变化过程与儿童近视性黄斑病变的进展密切相关。本课题研究发现我国高度近视儿童中近视性黄斑病变的发生率近50%，以豹纹状眼底最为常见。在近视性黄斑病变早期，各区域脉络膜厚度显著变薄，其中黄斑鼻侧的脉络膜厚度变化最为显著。通过构建多分类ROC模型，发现黄斑鼻侧脉络膜厚度是预测高度近视黄斑病变的良好指标。OCT是分析脉络膜三维结构的有效手段。基于课题组开发的深度学习算法，对OCT图像中的脉络膜血管成分进行增强、分割，计算得到脉络膜血管指数。随着近视度数加深、眼轴增长、脉络膜厚度变薄，脉络膜血管指数逐渐降低。这进一步说明随着近视度数的加深，脉络膜血管成分逐步减少。豹纹状眼底是近视性黄斑病变的早期阶段，在高度近视儿童中较为常见。基于课题组创立的豹纹状眼底分级标准，随儿童年龄增长、眼轴增长、近视度数加深，高等级豹纹状眼底的比例逐步增加，且脉络膜厚度均与豹纹状眼底分级呈负相关。为了进一步精确的分析豹纹状眼底，课题组运用深度学习方法提取、分割常规眼底彩照和超广角眼底照中的脉络膜血管成分，取得了良好的效果。基于上述深度学习方法提取的豹纹，结合眼底分区定量算法，获得眼底不同区域的豹纹状眼底密度。该分析手段作为对OCT脉络膜结构分析的补充，提供了一个新的视角和维度，是定量分析脉络膜三维结构的有效手段。综上，本课题从多个维度探索了高度近视儿童脉络膜结构改变及其相关因素，上述研究结果将为探索高度近视早期眼底改变的发生机制提供新思路，同时为高度近视的进展控制和眼底病变的预防与干预提供重要参考。上述研究成果在American journal of ophthalmology、Acta Ophthalmologica、Current Eye Research等杂志上共发表论著4篇。