Department of Ophthalmology, College of Medicine, Rangsit University, Rajavithi Hospital, Bangkok, Thailand.
Google Health, Palo Alto, CA, USA.
Lancet Digit Health. 2022 Apr;4(4):e235-e244. doi: 10.1016/S2589-7500(22)00017-6. Epub 2022 Mar 7.
Diabetic retinopathy is a leading cause of preventable blindness, especially in low-income and middle-income countries (LMICs). Deep-learning systems have the potential to enhance diabetic retinopathy screenings in these settings, yet prospective studies assessing their usability and performance are scarce.
We did a prospective interventional cohort study to evaluate the real-world performance and feasibility of deploying a deep-learning system into the health-care system of Thailand. Patients with diabetes and listed on the national diabetes registry, aged 18 years or older, able to have their fundus photograph taken for at least one eye, and due for screening as per the Thai Ministry of Public Health guidelines were eligible for inclusion. Eligible patients were screened with the deep-learning system at nine primary care sites under Thailand's national diabetic retinopathy screening programme. Patients with a previous diagnosis of diabetic macular oedema, severe non-proliferative diabetic retinopathy, or proliferative diabetic retinopathy; previous laser treatment of the retina or retinal surgery; other non-diabetic retinopathy eye disease requiring referral to an ophthalmologist; or inability to have fundus photograph taken of both eyes for any reason were excluded. Deep-learning system-based interpretations of patient fundus images and referral recommendations were provided in real time. As a safety mechanism, regional retina specialists over-read each image. Performance of the deep-learning system (accuracy, sensitivity, specificity, positive predictive value [PPV], and negative predictive value [NPV]) were measured against an adjudicated reference standard, provided by fellowship-trained retina specialists. This study is registered with the Thai national clinical trials registry, TCRT20190902002.
Between Dec 12, 2018, and March 29, 2020, 7940 patients were screened for inclusion. 7651 (96·3%) patients were eligible for study analysis, and 2412 (31·5%) patients were referred for diabetic retinopathy, diabetic macular oedema, ungradable images, or low visual acuity. For vision-threatening diabetic retinopathy, the deep-learning system had an accuracy of 94·7% (95% CI 93·0-96·2), sensitivity of 91·4% (87·1-95·0), and specificity of 95·4% (94·1-96·7). The retina specialist over-readers had an accuracy of 93·5 (91·7-95·0; p=0·17), a sensitivity of 84·8% (79·4-90·0; p=0·024), and specificity of 95·5% (94·1-96·7; p=0·98). The PPV for the deep-learning system was 79·2 (95% CI 73·8-84·3) compared with 75·6 (69·8-81·1) for the over-readers. The NPV for the deep-learning system was 95·5 (92·8-97·9) compared with 92·4 (89·3-95·5) for the over-readers.
A deep-learning system can deliver real-time diabetic retinopathy detection capability similar to retina specialists in community-based screening settings. Socioenvironmental factors and workflows must be taken into consideration when implementing a deep-learning system within a large-scale screening programme in LMICs.
Google and Rajavithi Hospital, Bangkok, Thailand.
For the Thai translation of the abstract see Supplementary Materials section.
糖尿病视网膜病变是可预防失明的主要原因，尤其是在低收入和中等收入国家（LMICs）。深度学习系统有可能增强这些环境中的糖尿病视网膜病变筛查，但评估其可用性和性能的前瞻性研究很少。
我们进行了一项前瞻性干预性队列研究，以评估将深度学习系统部署到泰国医疗保健系统中的实际性能和可行性。符合条件的患者为患有糖尿病且在国家糖尿病登记册中列出、年龄在 18 岁或以上、能够拍摄至少一只眼睛的眼底照片且根据泰国公共卫生部的指南需要进行筛查的患者。合格的患者在泰国国家糖尿病视网膜病变筛查计划下的九个初级保健点接受深度学习系统筛查。有糖尿病黄斑水肿、严重非增殖性糖尿病视网膜病变或增殖性糖尿病视网膜病变既往诊断史、视网膜激光治疗或视网膜手术既往史、需要转诊至眼科医生的其他非糖尿病视网膜病变眼病、或因任何原因无法拍摄双眼眼底照片的患者被排除在外。深度学习系统对患者眼底图像的解释和转诊建议实时提供。作为安全机制，区域视网膜专家对每张图像进行了重新阅读。深度学习系统的性能（准确性、敏感性、特异性、阳性预测值[PPV]和阴性预测值[NPV]）与由 fellowship-trained 视网膜专家提供的经裁决的参考标准进行了比较。本研究在泰国国家临床试验注册处（TCRT）注册，注册号为 TCRT20190902002。
2018 年 12 月 12 日至 2020 年 3 月 29 日期间，筛查了 7940 名患者以纳入研究。7651 名（96.3%）患者符合研究分析条件，2412 名（31.5%）患者因糖尿病视网膜病变、糖尿病黄斑水肿、无法评估图像或低视力而被转诊。对于威胁视力的糖尿病视网膜病变，深度学习系统的准确性为 94.7%（95%CI 93.0-96.2），敏感性为 91.4%（87.1-95.0），特异性为 95.4%（94.1-96.7）。视网膜专家重新阅读者的准确性为 93.5%（91.7-95.0；p=0.17），敏感性为 84.8%（79.4-90.0；p=0.024），特异性为 95.5%（94.1-96.7；p=0.98）。深度学习系统的 PPV 为 79.2%（95%CI 73.8-84.3），而重新阅读者的 PPV 为 75.6%（69.8-81.1）。深度学习系统的 NPV 为 95.5%（92.8-97.9），而重新阅读者的 NPV 为 92.4%（89.3-95.5）。
深度学习系统可以提供类似于社区筛查环境中视网膜专家的实时糖尿病视网膜病变检测能力。在中低收入国家实施大规模筛查计划时，必须考虑社会环境因素和工作流程。
Google 和泰国曼谷的 Rajavithi 医院。
