An Automated Diagnosis of Skin Cancer Disease Using Machine Learning Techniques

  • A.C.Kaladevi, P.Ramya, Elamathi S, Atchaya R, Dhanalakkshmi M

Abstract

The skin is the largest organ in the human body, which protects us from microbes and other pathogens. Etymologically, dermatology is the medical discipline in the analysis and prevention of skin abnormalities.  The system is an automated diagnostic system unlike the conventional system involving human arbitration based on the ideology of dermatological diagnosis. The system works in two dependent steps: the first detects skin abnormalities and the second identifies skin diseases. The system works with visual inputs, i.e. high-resolution color images and patient history. The automated diagnostic system uses a modified genetic algorithm, a k -means grouping, and an SVMclassifier to perform preprocessing segmentation, and featureextraction on the images respectively. To detect askin cancer disease, the system uses a neural network to propagate artificial feedback, which is implemented using MATLAB. The system has a skin cancer detection accuracy of 98.99% and a canceridentification accuracy of 97.016% when testing diseased areas on skin images. In addition to this, various systems have been proposed to assist researchers in the automatic detection of melanoma. This investigation focuseson thealgorithms usedfor the automated detection of melanoma in dermoscopicimages through a complete analysis of the stages of the proposed methodologies. It also examines the concepts associated with skin cancer disease and describes possible future directions through open problems in this area ofresearch.

 Keywords Dermatology, Melanoma, Images, Accuracy.

Published
2020-05-19
How to Cite
A.C.Kaladevi, P.Ramya, Elamathi S, Atchaya R, Dhanalakkshmi M. (2020). An Automated Diagnosis of Skin Cancer Disease Using Machine Learning Techniques. International Journal of Advanced Science and Technology, 29(06), 4115 - 4126. Retrieved from https://sersc.org/journals/index.php/IJAST/article/view/16417