Context: Doctors depend heavily on getting good images so that they can confidently diagnose challenging diseases and conditions. Thus, it's to make sure that the images that are captured are good (i.e. patient did not move during acquisition), in service of the doctors who will be interpreting the image. This is one challenge to which we apply machine learning!
Challenge The task is to build a model to predict the quality label of an image. Given images are of good (label 0), acceptable (label 1) and bad quality (label 2).
Data fundus folder consists of two components:
150 .pngs in the "fundus" folder consisting of fundus images of varying image quality a csv containing the image name and quality label