Please use this identifier to cite or link to this item:
Title: Skin lesion detection and recognition via deep learning
Authors: Chen, Ziyu
Keywords: Engineering::Electrical and electronic engineering
Issue Date: 2022
Publisher: Nanyang Technological University
Source: Chen, Z. (2022). Skin lesion detection and recognition via deep learning. Final Year Project (FYP), Nanyang Technological University, Singapore.
Project: P3040-202
Abstract: Melanoma, also known as malignant melanoma, is a type of cancer that develops from melanocytes, it is a rare form of skin cancer and arguably the most dangerous form of skin cancer. Melanoma is one of the leading causes of death due to its high degree of malignancy. Besides, some melanomas have a low contrast to the adjacent skin and are difficult for dermatologists to do physical detection and examination, so it is rather challenging to automatically apply segmentation techniques to them. This report proposes a medical image segmentation model and classification implementation model that will speed up the segmentation and diagnosis of melanoma. The image segmentation implementation of this project is constructed based on ISIC 2018 Task1 dataset that contains 2625 images. A network called U-Net variant based on a fully convolutional network (FCN) is used, which obtained 93.3% of Accuracy, 87.1% of Precision, 90.5% of Recall, 87.3% of F1, 85.1% Dice coefficient, and 79.3% of Jaccard. The image classification implementation of this project is constructed based on ISIC 2018 Task3 dataset that contains 10046 images. A Convolutional Neural Network (CNN) is used, which obtained an evaluation accuracy of 96.47%, specificity of 98.3%, the sensitivity of 90.3%, and precision of 92.2%.
Schools: School of Electrical and Electronic Engineering 
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
Chen Ziyu_FYP_Final_Report.pdf
  Restricted Access
2.25 MBAdobe PDFView/Open

Page view(s)

Updated on Dec 10, 2023


Updated on Dec 10, 2023

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.