Development of a deep learning model to detect crystalline retinopathy

This study was conducted using a publicly accessible FP database obtained from a previous study perfromed by Cen and additional anonymized FP data. Additional retinal image datasets were extracted from Google Images and Google search engine by searching for following keywords: crystalline retinopathy, talc retinopathy, and retinal crystal deposit. In this study, we aimed of building a deep learning model to detect crystalline retinopathy in FP images.

本研究使用了取自Cen此前开展的一项研究的公开可用FP（眼底摄影，Fundus Photography）数据库，以及额外的匿名化FP数据。额外的视网膜图像数据集则通过谷歌图片及谷歌搜索引擎，以"结晶样视网膜病变""滑石粉性视网膜病变""视网膜晶体沉积"为关键词检索获取。本研究旨在构建深度学习模型，以在FP图像中检测结晶样视网膜病变。