Ophthalmology Clinic, Department of Medicine and Ageing Science, University "G. D'Annunzio" of Chieti-Pescara, Via Dei Vestini Snc, 66100, Chieti, Italy.
Ophthalmology Unit, Department of Surgical, Medical, Molecular Pathology and Critical Care Medicine, University of Pisa, Via Roma 67, 56126, Pisa, Italy.
Graefes Arch Clin Exp Ophthalmol. 2024 Jan;262(1):149-160. doi: 10.1007/s00417-023-06170-6. Epub 2023 Aug 2.
To distinguish functioning from failed filtration blebs (FBs) implementing a deep learning (DL) model on slit-lamp images.
Retrospective, cross-sectional, multicenter study for development and validation of an artificial intelligence classification algorithm. The dataset consisted of 119 post-trabeculectomy FB images of whom we were aware of the surgical outcome. The ground truth labels were annotated and images splitted into three outcome classes: complete (C) or qualified success (Q), and failure (F). Images were prepared implementing various data cleaning and data transformations techniques. A set of DL models were trained using different ResNet architectures as the backbone. Transfer and ensemble learning were then applied to obtain a final combined model. Accuracy, sensitivity, specificity, area under the ROC curve, and area under the precision-recall curve were calculated to evaluate the final model. Kappa coefficient and P value on the accuracy measure were used to prove the statistical significance level.
The DL approach reached good results in unraveling FB functionality. Overall, the model accuracy reached a score of 74%, with a sensitivity of 74% and a specificity of 87%. The area under the ROC curve was 0.8, whereas the area under the precision-recall curve was 0.74. The P value was equal to 0.00307, and the Kappa coefficient was 0.58.
All considered metrics supported that the final DL model was able to discriminate functioning from failed FBs, with good accuracy. This approach could support clinicians in the patients' management after glaucoma surgery in absence of adjunctive clinical data.
利用深度学习（DL）模型在裂隙灯图像上区分功能性和失败性滤过泡（FB）。
这是一项回顾性、横断面、多中心研究，旨在开发和验证人工智能分类算法。该数据集包含 119 例青光眼滤过术后 FB 图像，我们对这些 FB 的手术结果有所了解。对这些图像进行了标注，并将其分为三种结局类别：完全（C）或合格成功（Q）和失败（F）。通过实施各种数据清理和数据转换技术来准备图像。使用不同的 ResNet 架构作为骨干网络训练了一组 DL 模型。然后应用迁移学习和集成学习来获得最终的组合模型。使用准确性、敏感度、特异性、ROC 曲线下面积和精度-召回曲线下面积来评估最终模型。准确性测量的 Kappa 系数和 P 值用于证明统计学显著性水平。
DL 方法在揭示 FB 功能方面取得了良好的效果。总体而言，该模型的准确率达到了 74%，敏感度为 74%，特异性为 87%。ROC 曲线下面积为 0.8，而精度-召回曲线下面积为 0.74。P 值等于 0.00307，Kappa 系数为 0.58。
所有考虑的指标都支持最终的 DL 模型能够区分功能性和失败性 FB，具有良好的准确性。这种方法可以在没有辅助临床数据的情况下，为青光眼手术后的患者管理提供支持。
