文献速递眼底病与葡萄膜炎

基于深度学习的地图样萎缩近红外反射成像评估

NEAR-INFRARED REFLECTANCE IMAGING FOR THE ASSESSMENT OF GEOGRAPHIC ATROPHY USING DEEP LEARNING.
PubMed 文献 · Retina · 2025-01-01

该研究在Rabin医疗中心开发并评估了全自动深度学习方法,用于在近红外反射图像上检测地图样萎缩(GA)。纳入113例GA患者和119例对照,分类模型准确率超过95%,Vision Transformer B16表现最佳(精确率98.5%、灵敏度98.4%、准确率98.5%);YOLOv8-Large定位灵敏度91%、精确率91%、IoU 84%、DICE系数88%。

要点

  • 近红外反射图像上可可靠识别地图样萎缩,深度学习模型分类准确率超过95%。
  • Vision Transformer B16分类表现最佳:精确率98.5%、灵敏度98.4%、准确率98.5%。
  • YOLOv8-Large定位表现:灵敏度91%、精确率91%、IoU 84%、DICE系数88%。

核心结论

  • 地图样萎缩可通过近红外反射图像可靠识别,深度学习模型可辅助评估该常规影像模态,帮助筛选可能受益于新兴疗法的患者。

研究设计与样本

  • 在Rabin医疗中心对50岁及以上、经两名视网膜专家确认的地图样萎缩患者的近红外反射图像进行分析。
  • 对照组纳入视网膜外观健康的患者的近红外反射图像。
  • 模型基于准确率、精确率、灵敏度、F1分数和DICE系数进行训练与评估。

主要发现

  • 共纳入113例地图样萎缩患者和119例对照。
  • 分类数据集包含330张图像,定位数据集包含659张图像。
  • 分类模型表现良好,准确率超过95%;Vision Transformer B16取得最佳结果(精确率98.5%、灵敏度98.4%、准确率98.5%)。
  • 地图样萎缩定位方面,YOLOv8-Large达到灵敏度91%、精确率91%、IoU 84%、DICE系数88%。

临床怎么用

  • 原文结论认为深度学习模型可辅助在近红外反射图像上评估地图样萎缩,帮助筛选可能受益于新兴疗法的患者。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

PURPOSE: Near-infrared reflectance imaging is a widely available but underused modality for assessing geographic atrophy (GA), a late-stage manifestation of dry age-related macular degeneration. The aim of this study was to develop and evaluate a fully automated deep-learning-based approach for detecting GA on near-infrared reflectance imaging. METHODS: Near-infrared reflectance images of patients aged 50 years or older with GA, confirmed by two retinal specialists, were analyzed at Rabin Medical Center. The control group included near-infrared reflectance images of patients with healthy-appearing retinas. Models were trained and evaluated based on accuracy, precision, sensitivity, F1 Score, and DICE coefficient. RESULTS: A total of 113 GA patients and 119 controls were included. The classification data set contained 330 images, and the localization data set included 659 images. Classification models performed well, with accuracy above 95%, while Vision Transformer B16 achieved the best results (precision = 98.5%, sensitivity = 98.4% and accuracy = 98.5%). For GA localization, YOLOv8-Large achieved 91% sensitivity, 91% precision, IoU of 84%, and DICE coefficient of 88%. CONCLUSION: Geographic atrophy can be reliably identified using near-infrared reflectance images. Deep learning models can assist in evaluating GA on this routinely available imaging modality, aiding in the selection of patients who may benefit from emerging therapies.

来源:Retina,查看原文。版权归原作者所有。

引用

  1. Fineberg A, Tiosano A, Golan N, et al. NEAR-INFRARED REFLECTANCE IMAGING FOR THE ASSESSMENT OF GEOGRAPHIC ATROPHY USING DEEP LEARNING. Retina. 2025. PMID: 40694826. doi:10.1097/IAE.0000000000004614

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。