3D Reconstruction and Analysis of Nevi for Melanoma Detection

Melanoma is the most aggressive skin cancer and its detection in the early stage is crucial for a successful medical treatment. Many algorithms have been developed to provide an objective examination, but none of them proved to be effective in clinical practice. We propose an innovative and automatic method that derives a 3D model of the patient body from multiple images through photogrammetry, identifies all nevi, and computes diagnostic indices related to asymmetry, border, color, diameter, and volume for each detected nevus. Good preliminary results were obtained from 15 subjects with both normal and melanoma suspicious nevi.