A Model for Recognizing Structureless Hyperpigmented Areas in Dermato-Oncology

A model for recognizing structureless hyperpigmented areas in images of skin neoplasms has been developed. Recognition of hyperpigmented areas is important for the diagnosis of skin melanoma, a rapidly progressing skin cancer. A digital dermatoscope RDS-2 has been used to obtain images serving as the initial data for the model. Software for recognizing hyperpigmentation areas in images of skin neoplasms has been developed on the basis of the proposed model. Tests have shown the recognition accuracy to be 82%. The proposed model can be recommended for use in decision-making support systems for the diagnosis of melanoma. © 2022, Springer Science+Business Media, LLC, part of Springer Nature.