Borrelli Enrico, Oakley Jonathan D, Iaccarino Giorgio, Russakoff Daniel B, Battista Marco, Grosso Domenico, Borghesan Federico, Barresi Costanza, Sacconi Riccardo, Bandello Francesco, Querques Giuseppe
Vita-Salute San Raffaele University Milan, Milan, Italy.
IRCCS San Raffaele Scientific Institute, Milan, Italy.
Eye (Lond). 2024 Feb;38(3):537-544. doi: 10.1038/s41433-023-02720-8. Epub 2023 Sep 5.
To validate a deep learning algorithm for automated intraretinal fluid (IRF), subretinal fluid (SRF) and neovascular pigment epithelium detachment (nPED) segmentations in neovascular age-related macular degeneration (nAMD).
In this IRB-approved study, optical coherence tomography (OCT) data from 50 patients (50 eyes) with exudative nAMD were retrospectively analysed. Two models, A1 and A2, were created based on gradings from two masked readers, R1 and R2. Area under the curve (AUC) values gauged detection performance, and quantification between readers and models was evaluated using Dice and correlation (R) coefficients.
The deep learning-based algorithms had high accuracies for all fluid types between all models and readers: per B-scan IRF AUCs were 0.953, 0.932, 0.990, 0.942 for comparisons A1-R1, A1-R2, A2-R1 and A2-R2, respectively; SRF AUCs were 0.984, 0.974, 0.987, 0.979; and nPED AUCs were 0.963, 0.969, 0.961 and 0.966. Similarly, the R coefficients for IRF were 0.973, 0.974, 0.889 and 0.973; SRF were 0.928, 0.964, 0.965 and 0.998; and nPED were 0.908, 0.952, 0.839 and 0.905. The Dice coefficients for IRF averaged 0.702, 0.667, 0.649 and 0.631; for SRF were 0.699, 0.651, 0.692 and 0.701; and for nPED were 0.636, 0.703, 0.719 and 0.775. In an inter-observer comparison between manual readers R1 and R2, the R coefficient was 0.968 for IRF, 0.960 for SRF, and 0.906 for nPED, with Dice coefficients of 0.692, 0.660 and 0.784 for the same features.
Our deep learning-based method applied on nAMD can segment critical OCT features with performance akin to manual grading.
验证一种深度学习算法用于对新生血管性年龄相关性黄斑变性（nAMD）中的视网膜内液（IRF）、视网膜下液（SRF）和新生血管性色素上皮脱离（nPED）进行自动分割。
在这项经机构审查委员会（IRB）批准的研究中，对50例（50只眼）渗出性nAMD患者的光学相干断层扫描（OCT）数据进行回顾性分析。基于两名盲法阅片者R1和R2的分级创建了两个模型A1和A2。曲线下面积（AUC）值衡量检测性能，并使用Dice系数和相关系数（R）评估阅片者与模型之间的定量关系。
基于深度学习的算法在所有模型和阅片者之间对所有液体类型均具有较高的准确性：每次B扫描中，IRF的AUC值在比较A1 - R1、A1 - R2、A2 - R1和A2 - R2时分别为0.953、0.932、0.990、0.942；SRF的AUC值分别为0.984、0.974、0.987、0.979；nPED的AUC值分别为0.963、0.969、0.961和0.966。同样，IRF的R系数分别为0.973、0.974、0.889和0.973；SRF的R系数分别为0.928、0.964、0.965和0.998；nPED的R系数分别为0.908、0.952、0.839和0.905。IRF的Dice系数平均为0.702、0.667、0.649和0.631；SRF的Dice系数分别为0.699、0.651、0.692和0.701；nPED的Dice系数分别为0.636、0.703、0.719和0.775。在手动阅片者R1和R2之间的观察者间比较中，IRF的R系数为0.968，SRF为0.960，nPED为0.906，相同特征的Dice系数分别为0.692、0.660和0.784。
我们应用于nAMD的基于深度学习的方法能够分割关键的OCT特征，其性能与手动分级相似。
