Tool to grade optic nerve head fundus images for disc damagelihood scale. DDLSNet takes the RDR and ARW values from RimNet,
combines them with a disc size category generated by a classification model (DiscNet), and calculates DDLS. The grading process is done in the script Final Pathway.py.
This repository contains the code used to train and evaluate the segmentation and classification models, but not the models themselves or the dataset used.
Running
Due to the way the repo is structured, if you wish to directly invoke a script that imports anything, you need to do python3 -m models.disc_size_model. See this Stack Overflow post.
This project requires TensorFlow >= 2.6.0.