Automated image-based biomarker computation tools for diabetic retinopathy

用于糖尿病视网膜病变的基于图像的自动化生物标志物计算工具

• 批准号: 10082344
• 负责人: Kaushal Solanki
• 金额: $ 30万
• 依托单位: EYENUK, INC.
• 依托单位国家: 美国
• 财政年份: 2012
• 资助国家: 美国
• 起止时间: 2012-09-01 至 2021-07-14
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10082344
• 关键词: Adult; Age; Appearance; Biological; Biological Markers; Biometry; Blindness; Caring; Clinical; Clinical Research; Code; Color; Communication; Computer Vision Systems; Consumption; Contracts; County; Detection; Devices; Diabetes Mellitus; Diabetic Retinopathy; Early identification; Environment; Evaluation; Exudate; Eye; Face; Faculty; Feedback; Fostering; Funding; Fundus; Goals; Health; Health Insurance Portability and Accountability Act; Health Services; Hemorrhage; Human; Image; Image Analysis; Individual; Industry; Institutes; Intervention; Intuition; Legal; Lesion; Los Angeles; Machine Learning; Market Research; Marketing; Measures; Microaneurysm; Monitor; Network-based; Ophthalmic examination and evaluation; Ophthalmology; Optometry; Participant; Patients; Pattern Recognition; Performance; Phase; Physicians; Pilot Projects; Price; Process; Prothrombin; Readiness; Research; Retina; Retinal Diseases; Risk; Sales; Small Business Innovation Research Grant; Small Business Technology Transfer Research; Surveys; System; Techniques; Testing; Time; Treatment Effectiveness; Uncertainty; Validation; Visualization software; Work; algorithm development; base; bioimaging; care providers; clinical care; cloud based; commercialization; computerized; computerized tools; convolutional neural network; design; diabetic patient; drug discovery; experience; fundus imaging; health economics; high risk; high throughput analysis; image registration; imaging biomarker; interest; large datasets; longitudinal analysis; medical schools; novel marker; novel therapeutics; payment; prevent; programs; prototype; retinal imaging; screening; serial imaging; sound; success; tool; usability

Abstract In this project, we present EyeMark, a system with advanced longitudinal image anal- ysis tools for automated computation of biomarkers for diabetic retinopathy (DR) using retinal fundus images. Specifically, we have developed tools for computation of microan- eurysm (MA) appearance and disappearance rates (jointly known as turnover rates) for use as a biomarker in quantifying DR progression risk along with longitudinal analysis of other DR lesions. The availability of a reliable image-based biomarker will have high pos- itive influence on various aspects of DR care, including screening, monitoring progres- sion, drug discovery and clinical research. Measuring MA turnover and longitudinal analysis of DR lesions involves two labor in- tensive steps: careful alignment of current and baseline images, and marking of individual lesions. This process is very time consuming and prone to error, if done entirely by human graders. The primary goal of this project is to overcome these limitations by automating both the steps involved in longitudinal analysis: accurate image registration, and lesion identification. We have designed and developed a prototype tool that robustly registers longitudinal images (even with multiple lesion changes) and effectively detects and localizes DR le- sions. This fully automated tool can work on the cloud to produces results in near constant time (for large datasets), and also provide intuitive visualization tools for clinicians to more effectively monitor DR progression. This commercialization readiness pilot (CRP) project is intended to develop a regulatory strategy and a market access plan for EyeMark to enable its introduction in the US market and foster commercial success.

抽象的 在这个项目中，我们展示了 EyeMark，一个具有先进纵向图像分析功能的系统 用于自动计算糖尿病视网膜病变 (DR) 生物标志物的分析工具 视网膜眼底图像。具体来说，我们开发了用于计算微电流的工具 eurysm (MA) 的出现率和消失率（统称为周转率） 用作量化 DR 进展风险的生物标志物以及纵向分析 其他 DR 病变。可靠的基于图像的生物标记物的可用性将具有很高的可能性 对 DR 护理的各个方面产生积极影响，包括筛查、监测进展 科学、药物发现和临床研究。 测量 MA 周转率和 DR 病变的纵向分析涉及两个劳动过程： 紧张的步骤：仔细对齐当前图像和基线图像，并对个体进行标记 病变。如果完全由人类完成，这个过程非常耗时并且容易出错 评分者。该项目的主要目标是通过自动化来克服这些限制 纵向分析涉及的两个步骤：准确的图像配准和病变 鉴别。 我们设计并开发了一个原型工具，可以稳健地记录纵向 图像（即使有多个病变变化）并有效检测和定位 DR le- 西翁。这个完全自动化的工具可以在云端工作，以近乎恒定的速度产生结果 时间（对于大型数据集），并且还为临床医生提供直观的可视化工具 有效监测 DR 进展。该商业化准备试点 (CRP) 项目 旨在为 EyeMark 制定监管策略和市场准入计划 使其进入美国市场并促进商业成功。