文献速递眼底病与葡萄膜炎

卢旺达基于人工智能的糖尿病视网膜病变筛查的可行性与接受度

Feasibility and acceptance of artificial intelligence-based diabetic retinopathy screening in Rwanda.
PubMed 文献 · Br J Ophthalmol · 2024-05-21

在卢旺达四家糖尿病诊所,2021年3月至6月对同意参与者用视网膜成像联合人工智能(AI)判读筛查糖尿病视网膜病变(DR),并与英国国家卫生系统认证的视网膜图像分级师判读对比;827名参与者中33.2%被AI转诊,AI对可转诊DR的敏感度92%、特异度85%,满意度99.5%,63.7%更偏好AI判读。

要点

  • AI筛查转诊率33.2%,其中单纯DR占32.0%、DR合并异常占39.6%、仅异常占23.6%、其他原因占4.73%。
  • 与人工判读相比,AI对可转诊DR的敏感度92%(95%CI 0.863%, 0.968%)、特异度85%(95%CI 0.751%, 0.882%)。
  • 参与者满意度99.5%,63.7%更偏好AI判读;DR转诊者随访依从性最高为53.4%。

核心结论

  • 在卢旺达糖尿病诊所使用AI进行DR筛查产生了准确的转诊,且参与者满意度高,提示AI筛查DR是实用且可接受的。

研究设计与样本

  • 2021年3月至2021年6月在卢旺达四家糖尿病诊所,对同意参与者使用视网膜成像联合AI判读筛查DR,并由英国国家卫生系统认证的视网膜图像分级师判读图像。

主要发现

  • 827名参与者(女性59.6%,n=493)接受AI筛查,其中33.2%(n=275)被转诊随访。
  • 与人工判读相比,AI对可转诊DR的敏感度为92%(95%CI 0.863%, 0.968%),特异度为85%(95%CI 0.751%, 0.882%)。
  • AI转诊者中,单纯DR为88例(32.0%),DR合并异常为109例(39.6%),仅异常为65例(23.6%),其他原因为13例(4.73%)。
  • 参与者对AI筛查满意度为99.5%(n=823),63.7%(n=527)更偏好AI而非人工判读。
  • 转诊依从性在因DR转诊者中最高,为53.4%。

关键数值与推荐

  • 基于国际糖尿病视网膜病变分类,DR分级≥2.0视为可转诊。
  • AI系统设计用于检测DR以外的视神经和黄斑异常;垂直杯盘比≥0.7和/或以60%及以上为切点识别的黄斑异常也被AI视为可转诊。

临床怎么用

  • 原文未据此给出推荐。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Evidence on the practical application of artificial intelligence (AI)-based diabetic retinopathy (DR) screening is needed. METHODS: Consented participants were screened for DR using retinal imaging with AI interpretation from March 2021 to June 2021 at four diabetes clinics in Rwanda. Additionally, images were graded by a UK National Health System-certified retinal image grader. DR grades based on the International Classification of Diabetic Retinopathy with a grade of 2.0 or higher were considered referable. The AI system was designed to detect optic nerve and macular anomalies outside of DR. A vertical cup to disc ratio of 0.7 and higher and/or macular anomalies recognised at a cut-off of 60% and higher were also considered referable by AI. RESULTS: Among 827 participants (59.6% women (n=493)) screened by AI, 33.2% (n=275) were referred for follow-up. Satisfaction with AI screening was high (99.5%, n=823), and 63.7% of participants (n=527) preferred AI over human grading. Compared with human grading, the sensitivity of the AI for referable DR was 92% (95% CI 0.863%, 0.968%), with a specificity of 85% (95% CI 0.751%, 0.882%). Of the participants referred by AI: 88 (32.0%) were for DR only, 109 (39.6%) for DR and an anomaly, 65 (23.6%) for an anomaly only and 13 (4.73%) for other reasons. Adherence to referrals was highest for those referred for DR at 53.4%. CONCLUSION: DR screening using AI led to accurate referrals from diabetes clinics in Rwanda and high rates of participant satisfaction, suggesting AI screening for DR is practical and acceptable.

来源:Br J Ophthalmol,查看原文。版权归原作者所有。

引用

  1. Whitestone N, Nkurikiye J, Patnaik JL, et al. Feasibility and acceptance of artificial intelligence-based diabetic retinopathy screening in Rwanda. Br J Ophthalmol. 2024. PMID: 37541766. doi:10.1136/bjo-2022-322683

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。