Welsh AI cancer trial delivers faster diagnosis in weeks, not months

TL;DR:

• Betsi Cadwaladr University Health Board in North Wales is using AI to triage cancer biopsies, cutting the time from initial sample to malignancy diagnosis from three months to one to two weeks.
• The trial uses the Paige PanCancer Detect platform and follows earlier successful pilots with prostate and breast cancer detection using the IBEX Galen system.
• Wales’s health secretary has backed the approach, saying AI is “able to find some of the problems that might otherwise be invisible.”

Wales’s largest health board, which covers the whole of North Wales and has some of the country’s worst cancer waiting times, is reporting early results from an AI pathology trial that dramatically compresses the gap between biopsy and diagnosis.

The problem the trial addresses is straightforward. Many biopsies taken for suspected cancer come back as benign and sit in laboratories for extended periods before being reviewed, depending on available resources. Audits showed that a significant number of these cases eventually turned malignant — meaning patients who appeared to be on a benign pathway were actually waiting months for a cancer diagnosis.

How the AI triage works

The Betsi Cadwaladr pilot uses the Paige PanCancer Detect platform, which analyses digitally scanned biopsy slides and flags potential malignancies for clinician review. Dr Muhammad Aslam, the national lead for digital pathology and AI projects, said the system creates a real-time triage that catches cases that would otherwise remain in a queue.

“Rather than those patients being given the diagnosis of malignancy after three months, they’ll get the diagnosis and the offer of treatment within one to two weeks,” he said.

The trial builds on earlier work. The health board previously ran prostate and breast cancer detection pilots using the IBEX Galen AI system, providing a track record that supported expanding to multi-cancer detection.

Political backing

Wales’s Health Secretary Jeremy Miles described the technology as a support tool rather than a replacement for clinicians. “It’s able to find some of the problems that might otherwise be invisible,” he said. The health board is seeking further funding to expand the trial before a potential national rollout.

Looking forward

AI-assisted pathology is not new in concept, but the speed gains reported here are notable. Compressing a three-month diagnostic window to two weeks is the kind of concrete, patient-facing improvement that makes the case for NHS AI investment in terms that matter beyond the technology itself. Whether the funding materialises for a Welsh national rollout will test how far early results translate into sustained commitment.