Retina OCT Angiography Images Classification

Purpose: Detect and differentiate active and inactive choroidal neovascularization (CNV) in different stages of age-related macular degeneration (AMD) using Optical Coherence Tomography Angiography (OCTA) scans. It is like a fine-grained classification problem since the categories belong to one disease but in different stages.

Challenges: The challenges are small dataset with unbalanced distribution and potential retina layer segmentation errors. It is also worthy noticing that currently, with only OCTA images, human cannot distinguish these categories very well either.