Multi-label classification of retinal disorders

Multi-label classification of retinal disorders

This is a deep learning course project. The goal is to develop a solution, using computer vision techniques, that can detect specific medical pathologies from patient's fundus images. The patient may have multiple pathologies.

The project involves several sub-tasks:

1. Build classification models
   • Using atleast two different model architectures
   • Explore transfer learning techniques
2. GradCAM - Visualize regions of interest that contribute to Diabetic Retinopathy and Glaucoma
3. Using the unlabeled dataset, augment the training data (semi-supervised learning) and report the change in classification performance on the (labeled) validation dataset

The notebook solution can be opened in google colab using the Open in Colab link inside the notebook.

Resources

• Retinal image dataset

Grad-CAM sample results