Sriwatana Kornchanok, Puttanawarut Chanon, Suwan Yanin, Achakulvisut Titipat
Department of Biomedical Engineering, Faculty of Engineering, Mahidol University, Nakhon Pathom, Thailand.
Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand.
Transl Vis Sci Technol. 2025 Jan 2;14(1):22. doi: 10.1167/tvst.14.1.22.
The purpose of this study was to develop a deep learning approach that restores artifact-laden optical coherence tomography (OCT) scans and predicts functional loss on the 24-2 Humphrey Visual Field (HVF) test.
This cross-sectional, retrospective study used 1674 visual field (VF)-OCT pairs from 951 eyes for training and 429 pairs from 345 eyes for testing. Peripapillary retinal nerve fiber layer (RNFL) thickness map artifacts were corrected using a generative diffusion model. Three convolutional neural networks and 2 transformer-based models were trained on original and artifact-corrected datasets to estimate 54 sensitivity thresholds of the 24-2 HVF test.
Predictive performances were calculated using root mean square error (RMSE) and mean absolute error (MAE), with explainability evaluated through GradCAM, attention maps, and dimensionality reduction techniques. The Distillation with No Labels (DINO) Vision Transformers (ViT) trained on artifact-corrected datasets achieved the highest accuracy (RMSE, 95% confidence interval [CI] = 4.44, 95% CI = 4.07, 4.82 decibel [dB], MAE = 3.46, 95% CI = 3.14, 3.79 dB), and the greatest interpretability, showing improvements of 0.15 dB in global RMSE and MAE (P < 0.05) compared to the performance on original maps. Feature maps and visualization tools indicate that artifacts compromise DINO-ViT's predictive ability but improve with artifact correction.
Combining self-supervised ViTs with generative artifact correction enhances the correlation between glaucomatous structures and functions.
Our approach offers a comprehensive tool for glaucoma management, facilitates the exploration of structure-function correlations in research, and underscores the importance of addressing artifacts in the clinical interpretation of OCT.
本研究旨在开发一种深度学习方法，用于恢复存在伪影的光学相干断层扫描（OCT）图像，并预测24-2 Humphrey视野（HVF）测试中的功能损失。
这项横断面回顾性研究使用了来自951只眼睛的1674对视野（VF）-OCT图像进行训练，以及来自345只眼睛的429对图像进行测试。使用生成扩散模型校正视乳头周围视网膜神经纤维层（RNFL）厚度图的伪影。在原始数据集和伪影校正后的数据集上训练了三个卷积神经网络和两个基于Transformer的模型，以估计24-2 HVF测试的54个敏感度阈值。
使用均方根误差（RMSE）和平均绝对误差（MAE）计算预测性能，并通过GradCAM、注意力图和降维技术评估可解释性。在伪影校正后的数据集上训练的无标签蒸馏（DINO）视觉Transformer（ViT）达到了最高的准确率（RMSE，95%置信区间[CI]=4.44，95%CI=4.07，4.82分贝[dB]，MAE=3.46，95%CI=3.14，3.79 dB），并且具有最大的可解释性，与原始图像上的性能相比，全局RMSE和MAE提高了0.15 dB（P<0.05）。特征图和可视化工具表明，伪影会损害DINO-ViT的预测能力，但通过伪影校正可以得到改善。
将自监督ViT与生成式伪影校正相结合可增强青光眼结构与功能之间的相关性。
我们的方法为青光眼管理提供了一个全面的工具，有助于在研究中探索结构-功能相关性，并强调了在OCT临床解释中解决伪影的重要性。
