The aim of this study is to develop a predictive model of risk of developing melanoma. The investigators will used artificial intelligence techniques to analysed images patterns obtained by clinical and dermoscopic pictures. The investigators want to defined a new predictive risk model obtained from genetic information and image analysis of pigmented lesions. This model could help to discriminate more accurately those patients who are most likely to develop melanoma in a high-risk population.

Inclusion Criteria:

• ≥18 years of age;
• Followed up patients in the Pigmented Lesions Unit at Reina Sofia University Hospital with high risk of developing melanoma (First-degree family/personal history of familiar melanoma and/or melanocyticnevi above 50 and/or dysplastic or atypical nevi);
• All nevi for which there is at least one dermatoscopy registry will be included.

Exclusion Criteria:

• not enough data in the patient clinical history;
• located lesions on the face, scalp, mucous membrane or on acral área;
• nodular or metastatic melanoma diagnosis
• type V or VI of Fitzpatrick phototyping scale

The patients were followed up in the Pigmented Lesions Unit at Reina Sofia University Hospital. In a subgroup of 200 patients, the investigators will analyze the dermoscopic imagines of melanocytic nevus, atypical nevus and melanomas to seek common patterns using artificial intelligence methods. With the patterns obtained, the investigators will build a robust model to classify population with the risk of developing melanoma. In this model the sociodemographic, clinical and genetic data will be used to increase the accuracy of the obtained model. In a second phase, in a subgroup of 100 different patients, the degree of accuracy of the predictive model will be evaluated.