Demographical, Morphological, and Histopathological Characteristics of Melanoma and Nevi Insights from Statistical Analysis and Machine Learning Models

Background: Early and accurate differentiation between melanomas and benign nevi is essential for making proper clinical decisions. This study evaluated 182 melanocytic le-sions using clinical, morphological, and histopathological parameters. Univariable anal-ysis was done in XLStat, while multivariable machine learning models were developed in Jamovi. Five supervised algorithms were compared, and glmnet (Elastic Net Regresion) was selected as best model due to superior balance of performance and calibration (AUC = 0.97). Age, horizontal diameter, and secondary histological features such as interactions with adjacent structures like epidermis (IAS-E) and changes in the extracellular matrix – like prominence of elastic fibers at the base of a lesion (CEM-BL) emerged as key predic-tors. While univariable and multivariable findings were consistent, machine learning al-lowed improved modeling of complex interactions. The results in our study demonstrate that structured demographyical, morphological and histopathological data can effectively support melanocitic lesions classification through machine learning approaches, offering a practical diagnostic support in dermatopathology.