Akter Nahida, Gordon Jack, Li Sherry, Poon Mikki, Perry Stuart, Fletcher John, Chan Thomas, White Andrew, Roy Maitreyee
School of Optometry and Vision Science, UNSW Sydney, Sydney, New South Wales, Australia.
School of Electrical and Data Engineering, University of Technology Sydney, Sydney, New South Wales, Australia.
PLoS One. 2025 Jan 17;20(1):e0316919. doi: 10.1371/journal.pone.0316919. eCollection 2025.
PURPOSE: In this study, we investigated the performance of deep learning (DL) models to differentiate between normal and glaucomatous visual fields (VFs) and classify glaucoma from early to the advanced stage to observe if the DL model can stage glaucoma as Mills criteria using only the pattern deviation (PD) plots. The DL model results were compared with a machine learning (ML) classifier trained on conventional VF parameters. METHODS: A total of 265 PD plots and 265 numerical datasets of Humphrey 24-2 VF images were collected from 119 normal and 146 glaucomatous eyes to train the DL models to classify the images into four groups: normal, early glaucoma, moderate glaucoma, and advanced glaucoma. The two popular pre-trained DL models: ResNet18 and VGG16, were used to train the PD images using five-fold cross-validation (CV) and observed the performance using balanced, pre-augmented data (n = 476 images), imbalanced original data (n = 265) and feature extraction. The trained images were further investigated using the Grad-CAM visualization technique. Moreover, four ML models were trained from the global indices: mean deviation (MD), pattern standard deviation (PSD) and visual field index (VFI), using five-fold CV to compare the classification performance with the DL model's result. RESULTS: The DL model, ResNet18 trained from balanced, pre-augmented PD images, achieved high accuracy in classifying the groups with an overall F1-score: 96.8%, precision: 97.0%, recall: 96.9%, and specificity: 99.0%. The highest F1 score was 87.8% for ResNet18 with the original dataset and 88.7% for VGG16 with feature extraction. The DL models successfully localized the affected VF loss in PD plots. Among the ML models, the random forest (RF) classifier performed best with an F1 score of 96%. CONCLUSION: The DL model trained from PD plots was promising in differentiating normal and glaucomatous groups and performed similarly to conventional global indices. Hence, the evidence-based DL model trained from PD images demonstrated that the DL model could stage glaucoma using only PD plots like Mills criteria. This automated DL model will assist clinicians in precision glaucoma detection and progression management during extensive glaucoma screening.
目的：在本研究中，我们调查了深度学习（DL）模型区分正常和青光眼视野（VF）以及将青光眼从早期到晚期进行分类的性能，以观察DL模型是否仅使用模式偏差（PD）图就能按照米尔斯标准对青光眼进行分期。将DL模型的结果与基于传统VF参数训练的机器学习（ML）分类器的结果进行比较。 方法：从119只正常眼睛和146只青光眼眼睛中收集了总共265张PD图和265个Humphrey 24-2 VF图像的数值数据集，用于训练DL模型，将图像分为四组：正常、早期青光眼、中度青光眼和晚期青光眼。使用两种流行的预训练DL模型：ResNet18和VGG16，通过五折交叉验证（CV）对PD图像进行训练，并使用平衡的、预增强的数据（n = 476张图像）、不平衡的原始数据（n = 265）和特征提取来观察性能。使用Grad-CAM可视化技术对训练后的图像进行进一步研究。此外，从全局指标：平均偏差（MD）、模式标准差（PSD）和视野指数（VFI）训练了四个ML模型，使用五折CV将分类性能与DL模型的结果进行比较。 结果：从平衡的、预增强的PD图像训练的DL模型ResNet18在对各组进行分类时取得了较高的准确率，总体F1分数为：96.8%，精确率：97.0%，召回率：96.9%，特异性：99.0%。对于原始数据集，ResNet18的最高F1分数为87.8%，对于采用特征提取的VGG16为88.7%。DL模型成功地在PD图中定位了受影响的VF损失。在ML模型中，随机森林（RF）分类器表现最佳，F1分数为96%。 结论：从PD图训练的DL模型在区分正常和青光眼组方面很有前景，并且与传统的全局指标表现相似。因此，从PD图像训练的基于证据的DL模型表明，DL模型仅使用PD图就能像米尔斯标准一样对青光眼进行分期。这种自动化的DL模型将有助于临床医生在广泛的青光眼筛查中进行精确的青光眼检测和病情进展管理。
Graefes Arch Clin Exp Ophthalmol. 2024-7
Ophthalmology. 2017-11-2
Front Med (Lausanne). 2022-3-17
Transl Vis Sci Technol. 2020-3
Zotero 插件
