A new study published in the British Journal of Dermatology by ACRF ACEMID researchers has developed an innovative way to measure sun damage across the entire body using artificial intelligence (AI) and 3D total-body photography.
Sun damage is the leading environmental risk factor for melanoma, but it is rarely included in risk assessment tools because there has been no simple and reliable way to measure it across the whole body. Current approaches often rely on people recalling past sun exposure, such as time spent outdoors or episodes of severe sunburn, which can be inaccurate.
To address this challenge, the ACEMID team developed a visual grading scale to assess sun damage in skin images. The study first tested the reliability of the scale and found strong agreement between clinicians and members of the public when rating the same images. Researchers then used community participants to label a large dataset of 24,750 skin image tiles collected from 107 volunteers who underwent 3D total-body photography. More than 200 image tiles were extracted from each participant's 3D digital avatar.
These labelled images were used to train an AI model to automatically detect sun damage. The model achieved high accuracy and performed even better when trained to recognise both sun damage and skin pigmentation. The researchers also validated the system using images from a separate group of participants, demonstrating its ability to perform well beyond the original dataset.
The AI presents results as colour-coded heatmaps on 3D body avatars (pictured below), highlighting areas of mild, moderate and severe sun damage. Overall, the study demonstrates that AI-powered 3D imaging can provide a fully automated measure of sun damage across the skin surface, offering a new tool to improve skin cancer risk assessment and support more personalised melanoma screening.
Link to publication: https://pubmed.ncbi.nlm.nih.gov/41370219/
