Betsi Cadwaladr: How AI is helping discover cancer for the first time in north Wales

  • Report by ITV Wales' North Wales reporter Ian Lang

A Welsh health board is leading the world with its use of artificial intelligence (AI) to help detect and diagnose cancer.

The team in Betsi Cadwaladr University Health Board is the first in the UK to pilot the use of the application, called the 'Galen platform', in a clinical setting to assist with breast cancer diagnosis.

Dr Muhammad Aslam believes this new clinical use could herald the dawning of a new era in cancer diagnostics, given that breast cancer treatment requires quick and accurate testing in order to get more successful outcomes for patients.

He said: “We are at the crossroads of an exciting new world and we are at the forefront within Betsi Cadwaladr.

“With the backlogs we have in the NHS it’s more important than ever to diagnose more quickly and more accurately.”

In Dr Aslam’s previous trial, using the application as a prostate cancer diagnostic tool, the programme looked at 1,600 biopsies and diagnosed 13% more cancers than without the AI help.

The real difference in this clinical setting is the clinicians are not only looking for the diagnostic cancer markers but also the prognosis markers, which will show which treatments or combinations of treatments will be effective for patients.

This could aid faster diagnosis of which specific cancers patients are suffering from, which in turn could inform treatment decisions more quickly.

Dr Anu Gunavardhan is the lead breast pathologist for BCUHB and the Quality Assurance Pathologist for Breast Test Wales. She is also chair of Welsh regional council of Royal College of Pathologists.

She said: “We need to get the results as quickly as possible, as there is lots of anxiety for the patients.

Pathologists will check the suspected results for accuracy and the system learns from their input for future tests. Credit: ITV Wales

“Our national targets are to get diagnoses back in three days and decide which treatment in seven days. We have prioritised breast cancers.

“We don’t always have enough staff to give this service, so we were looking at how this programme could help us.”

The Galen platform gains knowledge as it goes. Pathologists will check the suspected results for accuracy and the system learns from their input for future tests.

Dr Aslam believes we should call it “assistive intelligence” rather than artificial intelligence because a clinician will always be needed to give the final say on diagnoses.

In the prostate cancer tests it yielded 100% reliability.

When it indicated there were cancerous cells, there were, shortening the diagnostic process but, more importantly, giving assurance to pathologists they were catching as many cases as possible.

Want a quick and expert briefing on the biggest news stories? Listen to our latest podcasts to find out What You Need To Know...