Implementation of artificial intelligence for the detection of cutaneous melanoma within a primary care setting

This research article was an investigation of the prevalence and types of skin cancer in outdoor enthusiasts, namely people who like to exercise outdoors. It was completed as part of a multinational research team. We explored the effectiveness of using artificial intelligence (AI) to detect skin cancer, particularly malignant melanomas (MM). The study is especially relevant to Australia, which has the world's highest per capita incidence of MM.

We conducted a cross-sectional study that involved a survey and total body skin cancer screening using AI-embedded software. Participants included 423 individuals engaged in outdoor activities like surfing, swimming, and walking/running. The study found that the point prevalence for MM was highest among surfers (6.48%), followed by walkers/runners (4.3%), and swimmers (3.33%). These rates were significantly higher compared to the general Australian population.

Interestingly, the AI software used in the study did not perform as well as expected, showing a sensitivity of 53.33%, a specificity of 54.44%, and an overall accuracy of 54.17%. This indicates that while the potential of AI in detecting early stages of MM is promising, the particular version of the AI software used in this study might not be ready for widespread clinical adoption. However later the software was given an update.

The findings underscore the importance of sun-safe protection measures and regular skin screenings, especially for those who spend significant time outdoors.

The full research paper is open access and can be read in the scientific journal PeerJ here