Artificial Intelligence Mole-Mapping Surveillance Technologies for Clinicians: A Scoping Review (Preprint)

BACKGROUND Early detection of melanoma is critical with early-stage disease resulting in 97% five-year survival, whereas advanced melanoma can be fatal and costs up to 30-fold more to treat. Total body photography (TBP) helps clinicians monitor high-risk individuals who have multiple naevi (moles), but manual image comparison is time consuming and frequently exceeds capacity. AI mole-mapping technologies claim to automate lesion tracking and triage, yet their clinical readiness is uncertain. OBJECTIVE To map all clinician facing AI mole-mapping technologies and appraise the quality of evidence and regulatory status supporting their use in secondary care clinical settings. METHODS A scoping review followed the JBI manual for evidence synthesis. Searches of MEDLINE, Scopus, company and regulatory websites, grey literature, clinical-trial registers and Google identified available technologies that acquire total body photography images and apply automated lesion-matching, tracking or change-detection AI algorithms for adult surveillance, for use by clinicians. Data were extracted including technical features, regulatory approvals and all primary studies of supporting evidence. Evidence quality was rated against reporting checklists. RESULTS Seven commercially available AI mole-mapping platforms met inclusion criteria. Six use two-dimensional TBP; one produces a three-dimensional avatar. Only one system (ATBM master + Moleanalyzer Pro) holds the appropriate CE Class IIa regulatory mark; none had FDA clearance. Eighteen primary studies evaluated five systems; 2/7 (29%) platforms lacked any public evidence. Evidence included mostly diagnostic accuracy studies in design, involved modest sample sizes and predominantly comprised fair-skinned populations. Reported performance metrics were frequently high but inconsistent and few studies demonstrated workflow gains. No cost-effectiveness analyses were found. CONCLUSIONS AI-assisted TBP using commercial imaging platforms shows promising diagnostic accuracy and has the potential to reduce dermatology workload, but current evidence is sparse, methodologically heterogeneous and heavily skewed towards fair-skinned populations. Regulatory approvals lag behind commercial availability. Large, multi-centre trials across diverse skin types, transparent reporting and economic evaluations are needed before widespread adoption can be justified. CLINICALTRIAL N/A