An Integrated Software Platform for Accelerating Image-Driven Ophthalmic Research and Driving New Insights and Endpoints to the Clinic

项目摘要 Translational Imaging Innovations（TII）正在创建一个集成平台， 以及眼部图像和数据的共享。我们的Integrated Translational Imaging ™平台集成了电子 研究工作流管理系统，并提供了用于 控制图像和数据流向可视化和图像处理工具包。我们现在讨论一个具体的 眼科和视觉界的痛点：DICOM符合性。2021年，FDA正式 将DICOM视为眼科图像管理系统的共识标准， 研究所和美国眼科学会向眼科器械界发出呼吁， 采用DICOM。眼科DICOM标准由12个独立的标准组成。构建DICOM图像文件 要求a）将专有（或非DICOM）图像文件转换为DICOM图像对象，以及B）填充 带有元数据的DICOM标签。DICOM规定了所需标签和可选标签（根据标准，偶数标签） 并且允许非标准的专有标签（奇数编号的）。不兼容的图像格式和专有标签 在眼科，特别是OCT中广泛存在。因此，我们建议建立一个DICOM图像生成器 整合到TII ocuVault API中，并将DICOM渲染整合到TII图像管理应用程序ocuLink中。 将DICOM阅读和翻译功能添加到ocuLink和ocuVault解决了重大问题 为眼科研究和开发社区。我们提供了一种引擎，用于控制 图像返回给所有者。我们采用透明、协调的模型，确保DICOM标准工具 可以使用. DCM文件，同时在必要时维护专有信息。我们将DICOM模型应用于 不存在DICOM补充的图像模式，加速新技术的协调 在眼科学中不断发展。 我们的目标如下：1）实现DICOM渲染到ocuLink; 2）开发方法来翻译专有的 图像格式到DICOM; 3）开发ocuVault到DICOM元数据映射软件实用程序; 4）开发 任何到DICOM的导出方法; 5）为不具有DICOM的文件开发符合DICOM的元数据填充实用程序 嵌入式元数据;以及6）将ocuLink和ocuVault作为眼科成像提交FDA 510（k）许可 管理体系（FDA产品代码NFJ）。 -1-

Project Summary Translational Imaging Innovations (TII) is creating an integrated platform for the collection, curation, analysis and sharing of ocular images and data. Our Integrated Translational Imaging™ platform integrates an electronic research workflow management system with an image repository, or data lake, and provides methods for controlling the flow of images and data to a visualization and image processing toolkit. We now address a specific pain point within the Ophthalmology and Vision community: DICOM conformance. In 2021, the FDA officially recognized DICOM as a consensus standard for Ophthalmic Image Management Systems, and the National Eye Institute and American Academy of Ophthalmology issued a call to the ophthalmic device community to accelerate adoption of DICOM. DICOM for ophthalmology consists of 12 separate standards. Building a DICOM image file requires a) transforming a proprietary (or non-DICOM) image file into a DICOM image object, and b) populating DICOM tags with metadata. DICOM specifies required tags and optional tags (even number tags per the standard) and allows for non-standard proprietary tags (odd-numbered). Non-compliant image formats and proprietary tags are widespread in ophthalmology, particularly in OCT. We therefore propose to build a DICOM image generator into the TII ocuVault API and incorporate DICOM rendering into the TII image curation application, ocuLink. The addition of DICOM reading and translation capabilities into ocuLink and ocuVault solves significant problems for the ophthalmic research and development community. We provide an engine for providing control of the images back to the owner. We apply a transparent, harmonized model that ensures that DICOM standard tools can use .DCM files, while maintaining proprietary information where necessary. We apply the DICOM model to image modalities where a DICOM supplement does not exist, accelerating the harmonization of new technologies in constant development in ophthalmology. Our AIMS are as follows: 1) Implement DICOM rendering into ocuLink; 2) Develop methods to translate proprietary image formats to DICOM; 3) Develop an ocuVault – to – DICOM metadata mapping software utility; 4) Develop an any – to – DICOM export method; 5) Develop DICOM-conformant metadata fill utility for files that do not have embedded metadata; and 6) submit ocuLink and ocuVault for FDA 510(k) clearance as an Ophthalmic Imaging Management Systems (FDA Product Code NFJ). -1-