期刊文献+

免散瞳眼底照相联合人工智能糖尿病视网膜病变筛查系统在内分泌科的应用价值 被引量:3

Application value of non-mydriatic fundus photography combined with artificial intelligence diabetic retinopathy screening system in endocrinology department
原文传递
导出
摘要 目的 探讨免散瞳眼底照相联合人工智能(AI)DR筛查系统在内分泌科的应用价值。方法 选取2018年3月至2021年1月于河南科技大学第一附属医院内分泌科住院并参与DR筛查的DM患者2281例(4471眼)的眼底照片,对比AI与内分泌科医生阅片诊断结果符合率。随机抽取542例(1000眼)患者眼底照片由眼科医生阅片,以眼科医生的判定结果为金标准,分析内分泌科和眼科医生阅片诊断结果的一致性及AI与眼科医生阅片诊断结果的一致性。结果 AI系统和内分泌科医生阅片有无DR的阳性符合率88.72%,阴性符合率为78.25%,总符合率为81.57%。内分泌科与眼科医生阅片有无DR结果的一致性Kappa值为0.910,敏感度为93.63%,特异度为97.16%。AI与眼科医生阅片有无DR结果的一致性Kappa值为0.882,敏感度为84.85%,特异度为98.21%。结论AI DR筛查系统阅片与内分泌科医生及眼科医生诊断结果的一致性较高,具有较高的敏感度和特异度,可用于DR筛查。 Objective To explore the application value of non-mydriatic fundus photography combined with artificial intelligence(AI)diabetic retinopathy(DR)screening system in endocrinology department.Methods The fundus photos of 2281 patients(4471 cases)who were hospitalized in the Endocrinology Department of The First Affiliated Hospital of Henan University of Science and Technology from March 2018 to January 2021 and participated in DR screening were selected. The reading results from AI and endocrinologist were compared. The fundus photos of 542 patients(1000 cases)were randomly selected and read by ophthalmologists. Taking the judgment results of ophthalmologists as the gold standard,the diagnostic consistency between endocrinologists and ophthalmologists and AI were analyzed.Results Whether in patients with or without DR,the total coincidence rate of AI and endocrinologist was 81. 57%,with positive coincidence rate 88. 72% and the negative coincidence rate 78. 25%. The concordance Kappa value of endocrinologist and ophthalmologist was 0. 910,with the sensitivity 93. 63% and the specificity97. 16%. The concordance value of DR reading results between AI and ophthalmologist was 0. 882,with the sensitivity 84. 85% and the specificity 98. 21%.Conclusion The AI-DR screening system has a high diagnostic consistency with endocrinologists and ophthalmologists on DR screening.
作者 王曼丽 李博 张瑞瑞 贾竹敏 付留俊 刘婕 王巧丽 WANG Manli;LI Bo;ZHANG Ruirui(The First Afiliated Hospital of Henan University of Science and Technology,Luoyang 471003,Chinal)
出处 《中国糖尿病杂志》 CAS CSCD 北大核心 2022年第11期808-811,共4页 Chinese Journal of Diabetes
基金 2020年度河南省医学科技攻关计划项目(LHGJ20200593)。
关键词 免散瞳眼底照相 人工智能 糖尿病视网膜病变 筛查 No-mydriatic fundus photography Artificial intelligence Diabetic retinopathy Screening
  • 相关文献

参考文献4

二级参考文献81

共引文献5358

同被引文献29

引证文献3

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部