Diagnostic Innovations in Glaucoma Study (DIGS): Glaucoma and High Myopia

项目摘要 这项研究的目标是“青光眼的诊断创新研究(DIGS)：青光眼和高度近视”， 是克服高度近视患者开角型青光眼(OAG)检测的障碍 (MypOAG)。2010年，全球估计有14亿人患有近视，患病率为 到2050年，预计将迅速上升到47.5亿。此外，高度近视的人要多2.5倍 与无高度近视的患者相比，OAG患者更容易发生OAG。目前尚不清楚近视为什么会增加OAG的风险，但确实如此 可能至少部分与生物力学因素有关；近视眼较长的眼轴可能导致 筛板变形，Bruch膜的暂时移位，毛细血管旁改变和 血管因素；所有这些都会导致视神经对OAG损伤的易感性增加。考虑到更高的 近视眼视盘倾斜和乳头周围萎缩的患病率，结构和功能测试 通常，指导治疗决定的价值会降低。这项提案将提供必要的后续行动 建立以患者为中心检测高度近视患者OAG进展的最佳实践 人口。具体地说，这项提议将1)识别视神经头(ONH)3D形态参数 光学相干断层扫描(OCT)(分段和非分段)以区分近视 有和没有进行性OAG的眼睛；2)使用新的OCT特征(例如纹理)优化变化检测 和微血管系统)从单个ONH和黄斑扫描合并的宽视野(WFoV)图； 3)开发新的纵向和多模式深度学习模型来预测OAG进展。多数 重要的是，我们将提高对结构和结构之间复杂的时间关系的理解， 在不同的近视队列中，与年龄和OAG相关的功能和微血管改变。 具体地说，在具体目标1中(提高我们对ONH之间复杂关系的理解 衰老和OAG眼的形态和结构、功能和微血管的变化)，我们解决 有无近视眼ONH形态特征的几个假说 高度近视。我们假设ONH形态可以预测与年龄和OAG相关的结构， 功能和微血管的变化，它是快速进展的预测。具体目标2(改进 使用WFoV图、未分割的3D体积、ONH检测近视眼的OAG进展 形态学)，我们提出了几个假设，旨在使用新的DL检测和预测OAG进展 接近了。在具体目标3中，我们将建立一个基于云的数据管理和计算渠道， 将促进安全的DL模型开发和与VISION研究社区的广泛数据共享。

Project Summary The objective of this study, “Diagnostic Innovations in Glaucoma Study (DIGS): Glaucoma and High Myopia”, is to overcome barriers to the detection of open angle glaucoma (OAG) in individuals with high myopia (mypOAG). In 2010, there was an estimated 1.4 billion people worldwide with myopia and the prevalence is rapidly rising to an estimated 4.75 billion by 2050. Moreover, persons with high myopia are 2.5 times more likely to have OAG than those without high myopia. It is unclear why myopia increases the risk of OAG, but it is likely related at least in part to biomechanical factors; longer axial lengths in myopic eyes may result in deformation of the lamina cribrosa, temporal displacement of Bruch's membrane, parapapillary changes and vascular factors; these all lead to increased susceptibility of the optic nerve to OAG damage. Given the higher prevalence of tilted discs and peripapillary atrophy in myopic eyes, the structural and functional tests that usually guide treatment decisions are of diminished value. This proposal will provide essential follow-up to establish best practices for patient-centered detection of OAG progression in the challenging high myopia population. Specifically, this proposal will 1) identify optic nerve head (ONH) 3D morphologic parameters from optical coherence tomography (OCT) scans (segmented and unsegmented) to differentiate between myopia eyes with and without progressive OAG; 2) optimize change detection using novel OCT features (e.g. texture and microvasculature) from wide field of view (WFOV) maps merged from individual ONH and macula scans; and 3) develop novel longitudinal and multimodal deep learning (DL) models to predict OAG progression. Most importantly, we will improve our understanding of the complex temporal relationship between structural, functional and microvascular age- and OAG related changes in a diverse cohort across the range of myopia. Specifically, in Specific Aim 1 (To improve our understanding of the complex relationship between ONH morphology and structural, functional, and microvascular change in the aging and OAG eye), we address several hypotheses related to the characterization of myopic ONH morphology in healthy eyes with and without high myopia. We hypothesize that ONH morphology is predictive of age – and OAG related structural, functional and microvascular changes and that it is predictive of fast progression. In Specific Aim 2 (To improve detection of OAG progression in myopic eyes using WFOV maps, unsegmented 3D volumes, ONH morphology), we address several hypotheses designed to detect and predict OAG progression using novel DL approaches. In Specific Aim 3, we will establish a cloud-based pipeline for data curation and computation that will facilitate secure DL model development and extensive data sharing with the vision research community.