眼前节裂隙灯（白内障、翼状胬肉）数据集

该数据集源自爱尔大数据中心2024年3月1日至12月31日的裂隙灯眼前节影像数据，数据范围涵盖5个省份、5家医院，经专家标注后形成眼前节裂隙灯（白内障、翼状胬肉）数据集，数据规模：19580条。该数据集以临床应用为导向，基于晶状体点状混浊、车轮状混浊等 20 余种病灶特征进行标注，形成 AI 训练样本，供爱尔自研模型：眼前节图像质量评估与翼状胬肉识别模型、白内障筛查模型进行训练，通过眼前节图像特征识别和提取，识别白内障、后发障等多种眼前节疾病，输出裂隙灯影像分析报告，辅助临床实现眼前节病变早期识别与高效诊断。​

This dataset is derived from slit-lamp anterior segment imaging data collected by Aier Big Data Center from March 1 to December 31, 2024. It covers 5 provinces and 5 hospitals, and after professional expert annotation, it forms a slit-lamp anterior segment dataset focused on cataracts and pterygium, with a total of 19,580 samples. Clinically oriented, this dataset is annotated based on more than 20 lesion characteristics including punctate lens opacities and wheel-shaped lens opacities, serving as AI training samples. It is used to train Aier's self-developed models, namely the anterior segment image quality assessment and pterygium recognition model, as well as the cataract screening model. By identifying and extracting features from anterior segment images, these models can recognize various anterior segment diseases such as cataracts and posterior capsular opacification, generate slit-lamp image analysis reports, and assist clinicians in early identification and efficient diagnosis of anterior segment lesions.

搜集汇总

背景与挑战

背景概述

该数据集包含19580条眼前节裂隙灯影像数据，专用于白内障和翼状胬肉等眼部疾病的AI辅助诊断模型训练。数据经过专家标注，涵盖影像路径、质量、分类和分级信息，旨在提升临床诊断效率和准确性，支持远程诊疗和早期筛查应用。