The goal of this project is to perform a comparative study of state-of-the-art deep learning models for segmentation of mobile-acquired dermatological images. Particularly, the following application scenarios in dermatology involving handheld image acquisition will be explored:
1) Skin lesions (melanocytic and nonmelanocytic lesions);
2) Skin wounds (pressure ulcers, traumatic/surgical wounds and bruises).
Expand AICOS’ background knowledge in segmentation strategies based on deep learning, with expected direct impact in current and future Derma portfolio projects such as MpDS, MDevNet and DermAI.
Author: Catarina Andrade
Type: MSc thesis
Partner: FEUP – Faculdade de Engenharia da Universidade do Porto