Independent breakdown
SkinVision
Depends on your goalTechnology: AI image analysis of a single mole/lesion (smartphone camera) + skin-cancer risk algorithm
What it actually does technically
You photograph one specific mole or spot; a convolutional-neural-network algorithm returns a low/high risk rating and tells you whether to get it checked. Some markets add review by a dermatologist and lesion tracking over time.
It is a medical-risk triage tool aimed at skin cancer — not a cosmetic or routine-building app.
What peer-reviewed research says about accuracy
Independent evidence is weaker and messier than the company’s headline numbers. SkinVision cites studies reporting ~95% sensitivity; a 2020 BMJ systematic review pooled the evidence at roughly 80% sensitivity and 78% specificity and judged the studies small and low-quality.
A 2022 prospective study of 1,204 lesions found the app over-flagged many benign spots and that performance swung widely (sensitivity ~41–83%) depending on the reference standard used. These are human-lesion studies, but prone to selective recruitment.
Known limitations the company doesn't advertise
It can miss real melanomas (false negatives) and over-flag harmless spots (false positives), driving anxiety and unnecessary visits. Real-world accuracy by ordinary users is likely worse than in studies.
It assesses one spot at a time, struggles with hard-to-photograph areas, and its CE Class I mark is largely self-certified — not the same as rigorous diagnostic clearance.
Who funded the studies they cite
Several of the strongest accuracy figures come from studies funded or co-authored by the company, and independent reviewers have argued its reported sensitivity/specificity were likely overestimated by study design. It is a paid product, so there is incentive to emphasise favourable numbers.
Plain English verdict
A genuine medical-risk tool, not a cosmetic gimmick — but treat it strictly as a prompt to see a doctor, never a substitute for one. It is CE-marked in Europe and NOT FDA-cleared in the US, and independent research says it both misses some cancers and over-flags benign spots. Useful as an extra nudge to get a worrying mole examined; dangerous if a "low risk" result falsely reassures you.
Worth it if you want
- +People in the EU who want a low-cost nudge to get a changing mole professionally checked
- +Tracking a known spot over time alongside — never instead of — dermatologist visits
Not worth it if
- −You are in the US and expect an FDA-cleared diagnosis (it is not)
- −You would use a "low risk" result to avoid seeing a doctor about a worrying mole
- −You want cosmetic or routine analysis — that is not what this does