ARTIFICIAL INTELLIGENCE IN SKIN CANCER DETECTION: BENEFITS AND CHALLENGES FOR DERMATOLOGICAL PRACTICE

Abstract

This article sought to highlight the importance of artificial intelligence (AI) in dermatology, especially for detecting skin cancer. This is a literature review carried out using the PubMed and Scielo databases using the descriptors "artificial intelligence", "cancer", "skin" with the Boolean operator "AND", including articles published between 2019 and 2024, in Portuguese and English, available for free. The articles reported that there were benefits in the classification of skin lesions along with the analysis of the patient's clinical data (age, gender, location, and form of evolution of the lesion). Studies show that the average accuracy of AI algorithms applied to combined imaging modalities was 86%, similar to dermoscopy. With the advancement of AI technology, the effectiveness of skin image diagnosis increases and consequently there is an improvement in the prognosis. However, several challenges must be faced, such as the need for standardization in image acquisition and processing techniques and the creation of representative databases. It is essential to emphasize that automated detection of skin diseases should not replace clinical evaluation by doctors, but rather complement it.