AI Science Tracker
EN

Real-world post-deployment performance of a novel machine learning-based digital health technology for skin lesion assessment and suggestions for post-market surveillance.

Skin Analytics Ltd · 27 Apr 2023 · medicine

Discovery

Real-world post-deployment performance of a novel machine learning-based digital health technology for skin lesion assessment and suggestions for post-market surveillance.

Source check: Source checked. Not independent scientific confirmation.

Finding: In prospective use at two NHS sites, DERM versions A/B were assessed across 14,500 cases; among 8,571 lesions with confirmed outcomes, reported melanoma sensitivity was 95–100% and malignancy sensitivity 96–100%, while benign specificity was 40.7–73.4%. Manufacturer ties, selection and differential-verification limitations constrain interpretation.

Technology: The DERM deep-ensemble image classifier analyzed dermoscopic images and returned lesion classes and referral recommendations; the paper evaluates its deployed performance.

Publication: Preprint · Test context: prospective or independent check. Catalog categories, not an evidence grade.

AI contribution
DERM returned lesion-class and referral recommendations during deployment; the study assessed real-world diagnostic performance against available clinical outcomes.
Test and result
Among lesions with confirmed outcomes, the study reports melanoma sensitivity of 95–100%, malignancy sensitivity of 96–100%, benign specificity of 40.7–73.4%, and 15–31% eligible for discharge.
Limitation / still unknown
Prospective post-deployment pathways at University Hospitals Birmingham and West Suffolk (July 2021–October 2022), with histology/clinical references, second reads and repeat-presentation checks. Manufacturer employment/equity ties and selection and differential-verification limitations are reported.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://assets-eu.researchsquare.com/files/rs-2842280/v1/cee57512-b224-4dc2-8000-8fe7f195add5.pdf?c=1700579969

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: linked separately.

Review date: 22 Nov 2026. Review scope: Research Square v1 primary full text Methods/Results; version identity and first-posting date checked against primary version metadata..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 4/5 · Validation 4/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse medicine discoveries.