文献速递青光眼眼底病与葡萄膜炎

深度学习与青光眼专家:眼底照片中视盘特征对预测青光眼转诊的相对重要性

Deep Learning and Glaucoma Specialists: The Relative Importance of Optic Disc Features to Predict Glaucoma Referral in Fundus Photographs.
PubMed 文献 · Ophthalmology · 2019-09-24

研究开发并验证了一种深度学习算法,用于从彩色眼底图像预测可转诊的青光眼性视神经病变(GON)及视神经乳头(ONH)特征,并比较该算法与青光眼专家(GS)及其他眼保健提供者的性能。算法在三个数据集上的AUC分别为0.945、0.855和0.881,对可转诊GON的敏感性高于10名评分者中的7名(包括3名GS中的2名),特异性与多数评分者相当。GS和算法均认为垂直杯盘比≥0.7、神经视网膜边缘切迹、视网膜神经纤维层缺损和裸露的环线血管是与可转诊GON最相关的特征。

要点

  • 深度学习算法从眼底图像预测可转诊GON的AUC在三个数据集中分别为0.945、0.855和0.881。
  • 算法对可转诊GON的敏感性显著高于10名评分者中的7名(包括2名GS),特异性高于3名评分者(包括1名GS)。
  • GS和算法均认为垂直杯盘比≥0.7、神经视网膜边缘切迹、视网膜神经纤维层缺损和裸露的环线血管是最关键特征。

核心结论

  • 仅基于眼底图像训练的深度学习算法检测可转诊GON的敏感性高于眼保健提供者,特异性与之相当。
  • 该算法在基于完整青光眼检查作出诊断的独立数据集上保持良好性能。

研究设计与样本

  • 研究为算法的开发与验证。
  • 算法使用回顾性数据集中的86 618张图像训练,由43名评分者评估青光眼性ONH特征和可转诊GON。
  • 算法在三个数据集上验证:数据集A(1205张图像,每患者1张,18.1%可转诊),由GS专家组裁定;数据集B(9642张图像,每患者1张,9.2%可转诊),来自糖尿病远程视网膜筛查项目;数据集C(346张图像,每患者1张,81.7%可转诊),来自青光眼门诊。

主要发现

  • 算法预测可转诊GON的AUC在数据集A为0.945(95%CI 0.929-0.960),数据集B为0.855(95%CI 0.841-0.870),数据集C为0.881(95%CI 0.838-0.918)。
  • 算法预测青光眼性ONH特征的AUC范围为0.661至0.973。
  • 算法敏感性显著高于10名未参与确定参考标准的评分者中的7名(包括3名GS中的2名),特异性高于3名评分者(包括1名GS),其余相当。
  • GS和算法均认为与可转诊GON最相关的特征为:垂直杯盘比≥0.7、神经视网膜边缘切迹、视网膜神经纤维层缺损和裸露的环线血管。

关键数值与推荐

  • 可转诊GON定义为ONH外观可疑到足以支持转诊进行全面检查。
  • 算法在数据集A、B、C中预测可转诊GON的AUC分别为0.945(95%CI 0.929-0.960)、0.855(95%CI 0.841-0.870)、0.881(95%CI 0.838-0.918)。
  • 算法预测青光眼性ONH特征的AUC在0.661至0.973之间。
  • GS和算法均认为最关键特征为垂直杯盘比≥0.7、神经视网膜边缘切迹、视网膜神经纤维层缺损和裸露的环线血管。

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

PURPOSE: To develop and validate a deep learning (DL) algorithm that predicts referable glaucomatous optic neuropathy (GON) and optic nerve head (ONH) features from color fundus images, to determine the relative importance of these features in referral decisions by glaucoma specialists (GSs) and the algorithm, and to compare the performance of the algorithm with eye care providers. DESIGN: Development and validation of an algorithm. PARTICIPANTS: Fundus images from screening programs, studies, and a glaucoma clinic. METHODS: A DL algorithm was trained using a retrospective dataset of 86 618 images, assessed for glaucomatous ONH features and referable GON (defined as ONH appearance worrisome enough to justify referral for comprehensive examination) by 43 graders. The algorithm was validated using 3 datasets: dataset A (1205 images, 1 image/patient; 18.1% referable), images adjudicated by panels of GSs; dataset B (9642 images, 1 image/patient; 9.2% referable), images from a diabetic teleretinal screening program; and dataset C (346 images, 1 image/patient; 81.7% referable), images from a glaucoma clinic. MAIN OUTCOME MEASURES: The algorithm was evaluated using the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity for referable GON and glaucomatous ONH features. RESULTS: The algorithm's AUC for referable GON was 0.945 (95% confidence interval [CI], 0.929-0.960) in dataset A, 0.855 (95% CI, 0.841-0.870) in dataset B, and 0.881 (95% CI, 0.838-0.918) in dataset C. Algorithm AUCs ranged between 0.661 and 0.973 for glaucomatous ONH features. The algorithm showed significantly higher sensitivity than 7 of 10 graders not involved in determining the reference standard, including 2 of 3 GSs, and showed higher specificity than 3 graders (including 1 GS), while remaining comparable to others. For both GSs and the algorithm, the most crucial features related to referable GON were: presence of vertical cup-to-disc ratio of 0.7 or more, neuroretinal rim notching, retinal nerve fiber layer defect, and bared circumlinear vessels. CONCLUSIONS: A DL algorithm trained on fundus images alone can detect referable GON with higher sensitivity than and comparable specificity to eye care providers. The algorithm maintained good performance on an independent dataset with diagnoses based on a full glaucoma workup.

来源:Ophthalmology,查看原文。版权归原作者所有。

引用

  1. Phene S, Dunn RC, Hammel N, et al. Deep Learning and Glaucoma Specialists: The Relative Importance of Optic Disc Features to Predict Glaucoma Referral in Fundus Photographs. Ophthalmology. 2019. PMID: 31561879. doi:10.1016/j.ophtha.2019.07.024

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。