Three dimensional total body photography identifies cutaneous phenotypes associated with late-onset invasive melanoma risk

Comprehensive Summary

This study investigates whether three-dimensional total body photography (3D-TBP) combined with artificial intelligence can identify skin phenotypes associated with invasive melanoma risk in adults over 50 years old. The researchers analyzed 117 individuals over the age of 50 with a history of melanoma using 3D-TBP to count the number of site specific photodamage and naevus counts. They then applied an unsupervised clustering model (DBSCAN) to group patients by shared phenotypic patterns. The results showed that four phenotypic patterns were identified, with different combinations of photodamage and naevus count (with no group with severe photodamage and a high naevi count). Individuals with severe generalized photodamage and few naevi had the highest invasive melanoma risk, while those with moderate V-neck photodamage had few naevi and a lower risk. The authors suggest that the melanoma risk is not influenced by the total mole count or overall sun damage, but rather their spatial distributions across the body.

Outcomes and Implications

Traditional melanoma risk assessment often relies on overall mole counts or generalized photodamage without taking into consideration anatomical variability, which can lead to overlooking sites with high risk. AI-assisted 3D-TBP could enable clinicians to give more personalized surveillance plans, which can improve the early detection of invasive melanoma in patients over 50. These systems still need more validation, but show potential to be integrated into the melanoma monitoring system.