SCRS Conference Proceedings on Intelligent Systems
Eliminating racial bias at the time of detection Melanoma using Convolution Neural Network (CNN)
Authors: Md. Abdullah Al Noman Majumder, Eimon Hossain Taief, Md. Nurul Amin Bhuiyan, M.F. Mridha and Aloke Kumar Saha
Publishing Date: 25-04-2022
ISBN: 978-93-91842-08-6
Abstract
Melanoma considers deadly cancer that can cause the death of a person if not distinguished at an initial stage. Although Melanoma is most common in white skin people and can be detected at an early stage using AI, white skin people have a much lower death rate from this cancer. But when black skin people have Melanoma, AI can't detect it at an early stage because most of the time, the machines are trained with Dermoscopic pictures of white people, which leads to a higher mortality rate for black skin people. As a result, people don't want to trust the AI system at the time of Melanoma detection. In this paper, we proposed a model with whatever black or white skin it can easily detect using machine learning. In this case, we will use the Convolution Neural Network (CNN) of machine learning to detect Melanoma at an early stage so that the death rate caused by Melanoma cancer can be reduced. The proposed method can detect Melanoma with an accuracy of 88.9\% for both skin people which may significantly decrease the mortality rate.
Keywords
Melanoma, Machine Learning, CNN, Dermoscopic picture, Artificial Intelligence.
Cite as
Md. Abdullah Al Noman Majumder, Eimon Hossain Taief, Md. Nurul Amin Bhuiyan, M.F. Mridha and Aloke Kumar Saha, "Eliminating racial bias at the time of detection Melanoma using Convolution Neural Network (CNN)", In: Raju Pal and Praveen Kumar Shukla (eds), SCRS Conference Proceedings on Intelligent Systems, SCRS, India, 2022, pp. 271-285. https://doi.org/10.52458/978-93-91842-08-6-27