EyeStar-BIO: Software for Enhancing Digital Binocular Indirect Ophthalmoscopy Examinations

总结 拟议项目的目标是开发EyeStar-BIO，这是一套软件工具，旨在增强数字化 双目间接检眼镜检查（BIO），检查的最常见形式之一， 视网膜。使用BIO，眼科医生可以确定多种视网膜疾病的存在，例如糖尿病视网膜病变。 视网膜病、黄斑变性和早产儿视网膜病等。传统的BIO考试 不产生任何视觉文件。 EyeStar-BIO将具有目前未作为数字BIO考试的一部分提供的创新功能：（1） 自动图像质量评估，（2）自动拼接以创建视网膜的宽视场视图，（3） 纵向分析;（4）基于AI的疾病检测。该项目具有8亿美元的市场潜力， 检查仍然是眼科医生最常用的程序之一。 第一阶段项目的目标将通过三个具体目标实现。目标1：实施软件， 数字BIO视频帧的自动图像质量评估。目标2：开发图像拼接方法， 使用所述视频中的最佳质量帧来生成所述视网膜的宽视场视图。目标3：生成图像， 跟踪由于疾病进展或消退而导致的时间差异。 EyeStar-BIO软件将授权给数字BIO耳机制造商。未来的发展将 包括EyeStar-BIO的真实的实时实现和使用人工智能诊断视网膜疾病的能力。 智能

Summary The goal of the proposed project is to develop EyeStar-BIO, a suite of software tools designed to enhance digital binocular indirect ophthalmoscopy examinations (BIO), one of the most common forms of examination of the retina. Using BIO, an ophthalmologist can determine the presence of multiple retinal diseases such as diabetic retinopathy, macular degeneration, and retinopathy of prematurity, among others. The traditional BIO exam does not produce any visual documentation. EyeStar-BIO will feature innovative features that are not currently offered as part of digital BIO exams: (1) automatic image quality assessment, (2) automatic stitching to create a wide-field view of the retina, (3) longitudinal analysis, and (4) AI-based disease detection. This project has an $800M market potential as BIO examinations remain one of the most common procedures for ophthalmologists. The goals of this Phase I project will be accomplished via three specific aims. Aim 1: implement software for automatic image quality assessment of digital BIO video frames. Aim 2: develop image stitching methods to generate a wide field view of the retina using the best quality frames in the video. Aim 3: generate images to track differences in time due to disease progression or regression. The EyeStar-BIO software will be licensed to manufacturers of digital BIO headsets. Future developments will include real time implementations of EyeStar-BIO and the ability to diagnose retinal disease using artificial intelligence.