Performance of retinal fluid monitoring in OCT imaging by automated deep learning versus human expert grading in neovascular AMD.
Maximilian PawloffBianca S GerendasGabor Gyoergy DeakHrvoje BogunovicAnastasiia GruberUrsula Margarethe Schmidt-ErfurthPublished in: Eye (London, England) (2023)
Deep learning-based segmentation of retinal fluid performs reliably on OCT images. CSFT values are weak indicators for fluid activity in nAMD. Automated quantification of fluid types, highlight the potential of deep learning-based approaches to objectively monitor anti-VEGF therapy.
Keyphrases
- deep learning
- optical coherence tomography
- convolutional neural network
- diabetic retinopathy
- artificial intelligence
- endothelial cells
- machine learning
- optic nerve
- high resolution
- age related macular degeneration
- physical activity
- mesenchymal stem cells
- high throughput
- mass spectrometry
- climate change
- smoking cessation