The impact of patient clinical information on automated skin cancer detection

摘要: Skin cancer is one of the most common types around world. For this reason, over past years, different approaches have been proposed to assist detect it. Nonetheless, them are based only on dermoscopy images and do not take into account patient clinical information. In work, first, we present a new dataset that contains images, acquired from smartphones, information skin lesions. Next, introduce straightforward approach combine data using well-known deep learning models. These models applied presented combining with We comprehensive study show impact final predictions. The results obtained by both sets general improvement 7% in balanced accuracy for all addition, statistical test indicates significant differences between without considering data. achieved shows potential detection piece important leverage systems.