The study by Schmidt et al. (2025) explored a new method for classifying the state of a disease using medical images. They focused on neovascular age-related macular degeneration (nAMD) and developed a model that could interpret disease progression—whether it's getting better, worse, or staying the same—based on pairs of images taken over time. Their approach was unique in that it encoded images independently, was sensitive to the order of disease stages, and could adjust for uncertainty in the progression labels.
The researchers' model was particularly adept at learning from noisy data, which is common in real-world medical settings. They also found that their model could be applied to classify the activity of nAMD from single images with limited training data, demonstrating its potential for few-shot learning. This suggests that the model could be a valuable tool for diagnosing and monitoring nAMD, even when clear or large amounts of data are not available.
Clinical Trials
This is a list of upcoming or ongoing clinical trials that are actively recruiting and have been listed or updated in the last two weeks:
