Development, validation and assessment of an automated, topographic structure-function approach to the detection of glaucoma and its progression

项目摘要/摘要 没有单一的(试纸)测试来确认青光眼的存在或进展，青光眼是一组视神经疾病 神经病。因此，临床医生依靠结构和功能测试的特征变化来诊断 并对疾病进行监测。越来越多的证据挑战了“结构先于结构”的旧观念 功能“。研究表明，诸如基线患者条件、测试范例和测试等因素 措施可能会影响结构性和功能性青光眼损害的一致程度。研究表明， 报告说，通过视野(VF)测试评估的功能损害和结构损失，如 通过光学相干断层扫描(OCT)视网膜神经纤维层(RNFL)厚度的测量评估， 通常是重合的，通常是线性的。事实上，萨米斯博士和他的导师最近的工作 研究小组(胡德教授和德莫赖斯博士)提供了证据，证明一种广域OCT扫描的地形方法 基于R算法的24-2和10-2VF能够识别异常结构--异常功能 超过90%的眼睛与早期青光眼损害一致。这项提议的首要目标是 评估地形图结构-功能评价在青光眼诊断和青光眼检测中的能力 进步。在K99阶段，将改进工具算法，以提高其性能 在早期青光眼的鉴别中的作用。它还将扩展到包括其他类型的OCT扫描 协议(例如，圆圈扫描)，以及来自其他制造商的扫描，以扩展该方法的实用性。 同时，Tsamis博士将接受数据科学和统计方面的广泛培训，以增强 他的技能，支持他的职业发展，并提供了一条通往研究独立的道路。他将使用 这些技能在R00阶段开发了一种检测青光眼的地形图评估方法 早期青光眼和晚期青光眼的进展。将对新开发的技术进行评估 对照临床参考标准以及其他常规方法，这些方法可在 青光眼诊所。这项拟议的研究在多个层面上具有翻译潜力。首先，VF-OCT 地形评估应在评估的敏感性和准确性之间取得更好的平衡 比目前可用的方法进步的速度更快。这应该会减少患者就诊的频率。 和测试。其次，开发的R算法有可能提供一个标准化的工具，以进一步 提高对青光眼表型及其与青光眼发病关系的认识和评估 结构状态受临床主观性影响较小。最后，提出的方法可以加快速度。 青光眼临床试验(如神经保护)的设计和实施 建立更好的结构-功能终点，提供疾病及其 进步。这将导致研究人群规模和临床试验持续时间的减少，从而 他们的成本。

Project Summary / Abstract There is no single (litmus) test to confirm the presence or the progression of glaucoma, a group of optic neuropathies. Therefore, clinicians rely on characteristic changes on structural and functional tests to diagnose and monitor the disease. There is increasing evidence that challenge the old notion that “structure precedes function”. Research has shown that factors, such as baseline patient conditions, test paradigms, and test measures can affect the extent to which structural and functional glaucomatous damage agree. Studies have reported that functional damage, assessed by means of visual field (VF) testing, and structural loss, as evaluated by measures of optical coherence tomography (OCT) retinal nerve fiber layer (RNFL) thickness, typically coincide, often in a linear manner. As a matter of fact, recent work by Dr. Tsamis and his mentoring team (Prof Hood and Dr. De Moraes) provided evidence that a topographic approach of widefield OCT scans and 24-2 and 10-2 VFs, based on an R algorithm, can identify abnormal structure – abnormal function agreement in over 90% of eyes with early glaucomatous damage. The overarching goal of this proposal is to assess the capability of a topographic structure-function evaluation in glaucoma diagnosis and detection of progression. During the K99 phase, the tool-algorithm will be refined with the aim of improving its performance in the identification of early glaucomatous eyes. It will also be expanded to include other types of OCT scan protocols (e.g., circle scans), as well as scans from other manufacturers to extend the utility of this approach. At the same time, Dr. Tsamis will undergo extensive training in data science and statistics in order to enhance his skills, support his career development and provide a pathway towards research independency. He will use these skills during the R00 phase to develop a method of topographic assessment for detection of glaucoma progression in eyes with early and advanced glaucoma. The newly developed technique will be evaluated against a clinical reference standard as well as against other conventional methods, which are available in the glaucoma clinic. The proposed research has translational potential at multiple levels. First, the VF-OCT topographic assessment should provide a better balance between sensitivity and accuracy for the assessment of rates of progression than currently available methods. This should lead to reduced frequency of patient visits and testing. Second, the developed R algorithm has the potential to provide a standardized tool to further improve our understanding and assessment of glaucomatous phenotypes and their association with the structural status in a way less influenced by clinical subjectivity. Last, the proposed method could accelerate the design and conduction of clinical trials in glaucoma (e.g. neuroprotection) by demonstrating and establishing better structure-function endpoints, providing surrogate measures of the disease and its progression. This should lead to a reduction in study population size and duration of clinical trials, and thus their costs.