Model for Estimating the Heterogeneity of the Distribution of Globule Characteristics in Images of Skin Neoplasms

The problem of diagnosing skin melanoma by digital imaging of the tumor is considered. Clinical algorithms for detecting skin melanoma are briefly described. A review of works devoted to the automated assessment of distribution asymmetry of the shape, color, and area of globules – important signs of melanoma, is given. A model for assessing the distribution heterogeneity of globule characteristics on digital images in the diagnosis of skin neoplasms has been developed, and models of distribution heterogeneity indicators have been proposed. An experimental comparative assessment of indicator models was carried out using a software system developed in the C++ language. The most informative indicators of globule characteristics distribution heterogeneity have been determined. The maximum (93%) accuracy in assessing the distribution heterogeneity of globule characteristics was obtained for the indicator “reduced reciprocal of the highest frequency of occurrence of the measured areas of globules.” The results of the study can be useful in the development of medical decision support systems for the diagnosis of melanoma. © 2022, Springer Science+Business Media, LLC, part of Springer Nature.