Men Yevgeniy, Fhima Jonathan, Celi Leo Anthony, Ribeiro Lucas Zago, Nakayama Luis Filipe, Behar Joachim A
Andrew and Erna Viterbi Faculty of Electrical & Computer Engineering, Technion, Israel Institute of Technology, Haifa 3200003, Israel.
Faculty of Biomedical Engineering, Technion, Israel Institute of Technology, Haifa 3200003, Israel.
Physiol Meas. 2025 Jan 22;13(1). doi: 10.1088/1361-6579/ada86a.
. Diabetic retinopathy (DR) is a serious diabetes complication that can lead to vision loss, making timely identification crucial. Existing data-driven algorithms for DR staging from digital fundus images (DFIs) often struggle with generalization due to distribution shifts between training and target domains.. To address this, DRStageNet, a deep learning model, was developed using six public and independent datasets with 91 984 DFIs from diverse demographics. Five pretrained self-supervised vision transformers (ViTs) were benchmarked, with the best further trained using a multi-source domain (MSD) fine-tuning strategy.. DINOv2 showed a 27.4% improvement in L-Kappa versus other pretrained ViT. MSD fine-tuning improved performance in four of five target domains. The error analysis revealing 60% of errors due to incorrect labels, 77.5% of which were correctly classified by DRStageNet.. We developed DRStageNet, a DL model for DR, designed to accurately stage the condition while addressing the challenge of generalizing performance across target domains. The model and explainability heatmaps are available atwww.aimlab-technion.com/lirot-ai.
糖尿病视网膜病变（DR）是一种严重的糖尿病并发症，可导致视力丧失，因此及时识别至关重要。现有的基于数据驱动的从数字眼底图像（DFI）进行DR分期的算法，由于训练域和目标域之间的分布差异，往往难以实现泛化。为了解决这个问题，开发了深度学习模型DRStageNet，它使用了六个公共独立数据集，包含来自不同人群的91984张DFI。对五个预训练的自监督视觉Transformer（ViT）进行了基准测试，其中表现最佳的使用多源域（MSD）微调策略进行进一步训练。与其他预训练的ViT相比，DINOv2的L-Kappa提高了27.4%。MSD微调在五个目标域中的四个域中提高了性能。误差分析表明，60%的误差是由于标签错误导致的，其中77.5%被DRStageNet正确分类。我们开发了DRStageNet，这是一种用于DR的深度学习模型，旨在准确对病情进行分期，同时应对跨目标域泛化性能的挑战。该模型和可解释性热图可在www.aimlab-technion.com/lirot-ai上获取。
