The goal of this project is to apply Deep Learning Generative approaches to create new examples of images that can be used in the XAI component and data augmentation. The approved cases – synthetically generated images - can be used in the XAI component as a reference of normal or glaucomatous casesThe goal of this work is to study new forms of explainability, from classical features and synthetic data with ophthalmological meaning, integrated in a recent Computer-Aided Diagnosis system for Glaucoma, developed for edge computing.
Explore XAI approaches for TAMI (Transparent Artificial Medical Intelligence) project in the area of Glaucoma. Also, there is a demand to build decentralised screening tools for Glaucoma worldwide.
Author: Pedro Lopes
Type: MSc thesis
Partner: FEUP – Faculdade de Engenharia da Universidade do Porto