CaseStudies.Chat
← Back to the archive
change.archi2026 · tech

Bradford NHS clinic's AI triage lifted sessions from 24 to 32 patients

St Luke's Hospital adopted the DERM skin-lesion AI in April 2026: a third more patients per clinic, benign cases discharged in minutes, biopsies down about 10%.

What was changed

The dermatology team at St Luke's Hospital in Bradford adopted the DERM system — Deep Ensemble for the Recognition of Malignancy, built by Skin Analytics — in April 2026 for suspicious skin lesions, the BBC reported on October 5, 2026. A healthcare assistant takes three photographs of the mole or lesion and uploads them; the AI analyses the images in minutes. Patients with benign cases are given advice and discharged, while flagged cases go straight to a specialist in a one-stop clinic — and every image in Bradford is double-checked by a clinician.

The numbers: each session now handles 32 patients against 24 before the AI, and the trust says the technology has helped staff see one-third more patients in clinics. Unnecessary biopsies have dropped by about 10% since introduction. Skin Analytics claims the AI is 99.9% accurate at ruling out melanoma — a company figure — and the technology has been recommended by the National Institute for Clinical Excellence. Consultant plastic surgeon Zakir Shariff said it had 'a big impact on all the bottlenecks — right from waiting times for patients to unnecessary investigations'.

One patient showed the pattern: 73-year-old Laurence Patten, urgently referred by his GP, had a mole flagged as suspicious by the AI; the specialist's examination suggested it was likely benign, and the mole will be removed and tested to confirm. Associate specialist Nader Ghaderi said that without the AI pathway, 'the doctor would have done a biopsy or definitely a follow-up would have been arranged'.

Why it worked

Skin-lesion triage was the bottleneck; minutes-long AI assessment replaced slower first reviews.

Rapid benign discharge freed specialist time for the cases that needed it.

Clinician double-checks on every image kept the safety net while throughput rose.

A NICE recommendation gave the trust a vetted basis to adopt the pathway.

What can be applied

Put AI at triage and keep clinician double-checks: capacity comes from confident early discharge, not from skipping review.

Aftermath

The trust reports the higher session capacity and the biopsy reduction holding since the April introduction, with flagged patients still receiving specialist confirmation through the same-visit clinic.

Sources

  1. The clinic using AI to detect skin cancer earlier ↗