Bradford NHS deploys AI dermatology tool to triage skin cancer

TL;DR:

• Bradford Teaching Hospitals NHS Foundation Trust began using Skin Analytics’ Deep Ensemble for Recognition of Malignancy (DERM) AI on 30 April 2026 to triage skin-lesion referrals at St Luke’s Hospital, with a three-year deployment planned.
• The trust handles around 5,000 skin cancer referrals annually under the two-week pathway, of which roughly 8% (400 patients) are found to have malignant cancer. DERM is reported by Skin Analytics to be 99.7% accurate at ruling out skin cancers.
• Resultsense view: this is a concrete, measurable UK NHS deployment — exactly the kind of in-pathway AI use case the Department of Health and Social Care has been trying to scale. With 26 trusts now on DERM, Skin Analytics has quietly become one of the most-deployed clinical AI tools in the NHS.

The technology is in active use across Manchester University, Liverpool University Hospitals, University Hospitals Dorset and Dorset County Hospital NHS Foundation Trusts, among others. Bradford joins a deployment cohort that has been growing through 2024 and 2025.

How the pathway works

After GP referral, patients enter a new three-times-weekly tele-dermatology service. Healthcare staff photograph any suspicious skin lesions in detail. The DERM algorithm analyses each image for visual characteristics and produces a suspected diagnosis, directing the next step in the pathway.

If a lesion is flagged as suspicious, the patient is directed to the consultant dermatologist in a “one-stop clinic” located next door to the image-capturing clinic and running alongside it. The dermatologist sees patients whose mole is identified as cancerous and performs an immediate excision, sent to the laboratory for diagnosis.

Benign cases are redirected to non-urgent pathways, freeing two-week-pathway capacity for the higher-risk minority. Tom White, general manager for Musculoskeletal and Therapies, described the long-term ambition: “We will also have the capacity to see this service go out into the community and GP surgeries, which means that, in the future, patients won’t need to come to hospital.”

The numbers behind the case

The 5,000-referral / 8%-malignant breakdown at Bradford is a useful baseline. Most NHS dermatology departments report similar ratios — large numbers of two-week-pathway referrals where the majority are benign. This is precisely the kind of high-volume, low-prevalence triage where image-classification AI can deliver measurable productivity gains, by routing benign cases out of the consultant queue without adding clinical risk.

Zakir Shariff, consultant plastic surgeon and clinical lead for skin cancer at the trust, called DERM “the future of skin cancer diagnosis in this country” and said it would help doctors “concentrate on treating the most urgent cases”.

UK relevance

For NHS England, this lands as a useful counter-narrative the same week the organisation has had to lock down hundreds of public GitHub repos over Mythos cyber-risk concerns. AI in the NHS is, in practice, doing both at once: defensive consolidation against frontier-model risk in IT, and active deployment in clinical pathways.

For UK SMEs in clinical AI, Skin Analytics’ trajectory — building a 26-trust footprint by stacking proven, in-pathway pilots — is the deployable template. The pattern (single high-volume specialty, measurable accuracy claim, fits inside the existing NHS pathway, redirects clinician time rather than replaces it) is repeatable in radiology, ophthalmology and pathology.

Looking forward

The next data point is the trust’s first published outcome data — typically 6–12 months post-deployment for AI tools in this category, covering volumes triaged, sensitivity and specificity in real-world use, and clinician time saved. UK clinical-AI vendors should expect Bradford’s published outcomes to influence procurement conversations across mid-sized trusts in 2026/27.