Kinder Scott, McNamara Steve, Clark Christopher, Bearce Benjamin, Thakuria Upasana, Veturi Yoga Advaith, Deitz Galia, de Carlo Forest Talisa E, Mandava Naresh, Kahook Malik Y, Singh Praveer, Kalpathy-Cramer Jayashree
Department of Ophthalmology, University of Colorado Anschutz Medical Campus, Aurora, CO, USA.
Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, USA.
Transl Vis Sci Technol. 2025 Jun 2;14(6):30. doi: 10.1167/tvst.14.6.30.
To develop an artificial intelligence (AI) optic cup and disc segmentation pipeline for obtaining optic nerve head (ONH) measurements such as vertical cup-to-disc ratio (VCDR) from fundus images and externally validate performance against optical coherence tomography (OCT) measurements.
This diagnostic study used a retrospectively collected dataset of 27,252 fundus images associated with 12,477 OCT reports and 21,714 expert assessments of VCDR from electronic health records (EHRs) for 4289 patients inclusive of glaucoma suspects, primary and secondary glaucoma. The AI pipeline was trained on nine public glaucoma datasets and externally validated on a private hospital dataset and a publicly available dataset.
AI VCDR predictions against OCT yielded mean absolute error (MAE), Pearson's R, and concordance correlation coefficient (CCC) values of 0.097 (95% confidence interval [CI], 0.095-0.099), 0.80 (95% CI, 0.79-0.81), and 0.66 (95% CI, 0.64-0.67), respectively. EHR VCDRs against OCT had MAE, Pearson's R, and CCC values of 0.086 (95% CI, 0.084-0.087), 0.77 (95% CI, 0.76-0.78), and 0.74 (95% CI, 0.73-0.75), respectively. The coefficient of variation (CV) of the AI pipeline on same-day images was 2.79%.
The proposed AI pipeline had strong correlation with OCT measurements and performed comparably to EHR assessments, with high repeatability. Increased diversity and cardinality of training data improved performance and generalizability to unseen datasets.
AI pipelines for fundus images can provide ONH measurements such as VCDR near expert level in new patient populations without the need for additional model training.
开发一种人工智能（AI）视杯和视盘分割流程，用于从眼底图像中获取视神经乳头（ONH）测量值，如垂直杯盘比（VCDR），并针对光学相干断层扫描（OCT）测量值进行外部性能验证。
这项诊断性研究使用了一个回顾性收集的数据集，包含27252张眼底图像，这些图像与12477份OCT报告以及来自4289名患者（包括青光眼疑似患者、原发性和继发性青光眼患者）电子健康记录（EHR）中的21714次VCDR专家评估相关。该AI流程在九个公开的青光眼数据集上进行训练，并在一家私立医院数据集和一个公开可用数据集上进行外部验证。
AI针对OCT的VCDR预测得出的平均绝对误差（MAE）、皮尔逊相关系数（Pearson's R）和一致性相关系数（CCC）值分别为0.097（95%置信区间[CI]，0.095 - 0.099）、0.80（95% CI，0.79 - 0.81）和0.66（95% CI，0.64 - 0.67）。EHR针对OCT的VCDR的MAE、Pearson's R和CCC值分别为0.086（95% CI，0.084 - 0.087）、0.77（95% CI，0.76 - 0.78）和0.74（95% CI，0.73 - 0.75）。该AI流程在同日图像上的变异系数（CV）为2.79%。
所提出的AI流程与OCT测量值具有强相关性，并且与EHR评估表现相当，具有高重复性。训练数据的多样性和基数增加可提高性能以及对未见数据集的泛化能力。
用于眼底图像的AI流程可以在新患者群体中提供如VCDR等接近专家水平的ONH测量值，而无需额外的模型训练。
