The incidence of malignant melanoma has significantly increased in the last four decades. Dermatologists are rarely present in rural or remote areas to perform an early detection of malignant melanoma. Our contribution is a low cost software that automatically and objectively differentiates between a melanoma lesion and a benign nevus in a simple, noninvasive manner. Our approach is based on the “ABCDE” classification of lesions, image processing, and artificial neural networks. The software was developed using images of previously diagnosed malignant melanomas and non-malignant suspicious moles, obtaining a sensibility of 76.56% and a specificity of 87.58%.
CITATION STYLE
Marín, C., Alférez, G. H., Córdova, J., & González, V. (2015). Detection of melanoma through image recognition and artificial neural networks. In IFMBE Proceedings (Vol. 51, pp. 832–835). Springer Verlag. https://doi.org/10.1007/978-3-319-19387-8_204
Mendeley helps you to discover research relevant for your work.