Classification of diabetes-related retinal diseases using a deep learning approach in optical coherence tomography.
Computer Methods and Programs in Biomedicine(2019)
摘要
•An end-to-end deep learning-based method for automatic classification of B-scans inside a volume for three retinal diseases.•The proposed model includes a feedback stage that highlights the areas of the scans to support the interpretation of the results. This information is potentially useful for a medical specialist while assessing the prediction produced by the model.•The proposed model tested on SERI+CUHK data set with healthy, DME and DR-DME patients obtained a precision of 0,93 and an AUC of 0,86. On the A2A SD-OCT data set the model outperformed the state-of-the-art methods with an AUC of 0.99 for AMD diagnosis.
更多查看译文
关键词
Optical coherence tomography,Deep learning models,Interpretability,Retinal diseases,Medical findings
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络