A study from Sweden shows that artificial intelligence can identify individuals at high risk of melanoma up to five years before the disease appears, improving prevention and diagnosis.

A new study conducted by researchers at the University of Gothenburg, Sweden, suggests that artificial intelligence (AI) can identify individuals at high risk of melanoma up to five years before the disease appears. Analyzing medical data from approximately 6 million adults, the researchers used AI models that included complex information, such as medical history and sociodemographic data, unlike traditional methods that focus on age and sex. The results showed an accuracy of 73% in identifying melanoma cases, compared to 64% in classical assessments. The study highlighted the ability to identify small groups at extremely high risk, where the probability of developing melanoma in the next five years is about 33%. This paves the way for effective preventive interventions. Melanoma, an aggressive form of skin cancer, requires early detection to improve survival rates. The authors of the study suggest using AI to prioritize patients based on risk, but emphasize the need for further studies and policy decisions to integrate these technologies into medical practice.