Skin lesion (melanoma) segmentation

• Main Author: Ong, Shi Quan
• Other Authors: Jiang Xudong
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence; Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/149427
• Institution: Nanyang Technological University

Skin lesions can pose serious health problems if left undetected and untreated. There are detection methods such as excisional biopsy as well as dermoscopy. However, there has been a demand of automating and aiding doctors with their diagnosis through Artificial Intelligence. The aim of this project is to evaluate the effectiveness of a Deep Learning model, U-NET, in identifying and segmenting the skin lesions from the ISIC 2017 Challenge Dataset. The evaluation consists of testing the U-NET model’s predicted segmentation accuracy with respect to different image and lesion types. Through this project, it can be concluded that the U-NET is effective in producing accurate segmentations of skin lesions.