Skin cancer detection : spreading awareness through mobile application

• Main Author: Christianti, Gabriella Benedicta
• Other Authors: Owen Noel Newton Fernando
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Computer science and engineering
• Online Access: https://hdl.handle.net/10356/148182
• Institution: Nanyang Technological University

Skin cancer is a growing danger, especially the elderly population. Early detection and treatment is critical in reducing the fatality of melanoma. Screening done by a specialist or individual can lead to improved survival rate of melanoma as melanoma is detected at reduced thickness and is easier to treat. However, only a small fraction of individuals have performed a full body skin examination at least once in their lives as they lack education on their risks of developing skin cancer. This project aims to spread awareness on individual’s risk of developing skin cancer through four assessments: Skin Cancer Detection Algorithm, Skin Risk Assessment, Profile Risk Assessment, and Lesion Risk Assessments. With increased awareness, the general public is expected to be more cautious of suspicious skin lesions. This increase in caution is expected to encourage the general public to go for a check-up with a specialist earlier, thus facilitating early detection and treatment of skin cancer and reducing the mortality of skin cancer. An ensemble machine learning model is created by combining InceptionResNetV2 and DenseNet201 trained on the HAM10000 dataset. This ensemble model achieved an average macro-average accuracy of 86% and is used as the Skin Cancer Detection Algorithm. The main findings of the study are as follows: (1) The users find the application useful and easy to use, (2) The users display positive attitude and behavioural intention towards using the application, and (3) The usefulness and usability of the application both significantly positively influenced the Attitude and Behavioural Intention towards using the application.