five

Red reflex examination for early detection of pediatric cataract using a novel red-reflex screening device

收藏
NIAID Data Ecosystem2026-05-10 收录
下载链接:
https://figshare.com/articles/dataset/Red_reflex_examination_for_early_detection_of_pediatric_cataract_using_a_novel_red-reflex_screening_device/31817661
下载链接
链接失效反馈
官方服务:
资源简介:
Early detection of pediatric cataracts is essential to prevent irreversible visual impairment; however, effective screening remains limited in many settings. This study evaluated the diagnostic performance of the CatScreener, a self-designed handheld red-reflex screening device, when operated by non-expert examiners. A total of 250 eyes from 250 infants and young children (≤8 years) underwent CatScreener screening and slit-lamp examination. Images were independently interpreted by two medical students and two junior ophthalmologists. Diagnostic outcomes were compared with an expert-defined reference standard. Cumulative summation (CUSUM) analysis was used to assess the learning curve for operational proficiency. Diagnostic accuracy did not differ significantly between CatScreener screening and slit-lamp examination across all examiners (all p > 0.05). The CatScreener demonstrated high sensitivity (84.9%-87.1%) and specificity (89.2%-91.0%), with excellent agreement with slit-lamp examination (Kappa = 0.856–0.952) and good inter-examiner consistency (Kappa = 0.792–0.920). Sensitivity and specificity remained stable across chronological operation stages and age-stratified subgroups. CUSUM analysis indicated that approximately 50 cases were required to achieve operational proficiency. The CatScreener is an effective and user-friendly tool for the early detection of pediatric cataracts, demonstrating high diagnostic accuracy, strong agreement with conventional examination methods, and consistent performance across examiners with varying levels of clinical experience.
创建时间:
2026-03-20
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作