AI reads chest X-rays first in north east London. Urgent results at one trust came back in 5 days instead of 13.

LONDON, November 2025 — North east London's cancer alliance has published early results from putting AI in front of radiologists on chest X-rays. Here's what it found, what it cost, and what went wrong.

Cost and outcome (as of Nov. 28, 2025)

Most urgent chest X-rays at Barts Health: turnaround cut 61.5%, from 13 days to 5. (NELCA)
Urgent cancer cases: reporting time cut 41.7%, from 12 days to 7. (NELCA)
Target: a consistent turnaround of under 3 days for all chest X-ray reports, not yet met. (NELCA)
Funding: £1m over two years. (NELCA)

The problem

Demand for chest X-rays in north east London rose 94% in five years, from 139,080 in 2017 to about 270,000 in 2022, the alliance says. It cites a 29% national shortfall in radiologists, and says some patients waited up to 21 days for results.

When it launched the project in November 2024, the alliance set a goal of cutting the wait for chest X-ray results "from three weeks to just three days for scans with significant findings." The alliance serves about 3 million people across eight boroughs. Its funding page projected that radiology demand in the area would grow by around 20% over the next decade, and chest X-ray demand by 29%.

How it works

Qure.ai's qXR reads each chest X-ray and flags those with findings that could indicate lung cancer, so radiologists and reporting radiographers see them first. It also sorts out normal scans quickly. The alliance says the technology is used "to assist, rather than replace," specialist teams. It runs through Sectra's Amplifier Services across the three trusts.

When the alliance won its £1m, it said it was one of only nine networks nationally to receive funding in full. The project follows the National Optimal Lung Cancer Pathway guidelines, it said. "This provides patients with faster access to crucial diagnostic tests," said Oliver Hawkins, the alliance's senior project manager, at launch.

What the alliance reported

At Barts Health, the most urgent chest X-rays now come back in 5 days rather than 13, and urgent cancer cases in 7 rather than 12. "The early results from using AI in chest X-rays are encouraging," said Wayne Douglas, the alliance's programme lead for diagnosis and treatment. "We are beginning to reduce reporting delays, improve prioritisation, and provide greater assurance for patients and clinicians."

Bhargava Reddy, Qure.ai's chief business officer for oncology, said the results show "how AI can play a supportive role in diagnostic pathways." The alliance says clinical teams gave positive feedback and that junior doctors and radiographers are being trained on the tool. A two-year evaluation is underway, with early findings shared with NHS England, and the alliance plans to extend AI to CT scanning.

What went wrong

One trust's figures. The release gives numbers only for Barts Health, not for BHRUT or Homerton.
Still short of the goal. Five days for the most urgent scans is better than 13, but above the alliance's own target of under three.
Reporting time, not diagnosis. The figures measure how fast X-rays are reported, not how fast patients are diagnosed or treated. The alliance says it will track time to CT and to diagnosis.
Not independently checked. The vendor is quoted in the alliance's release and republished the figures; we found no outside audit of them.
What we don't know yet: the two-year evaluation's results, and what the tool costs once the £1m runs out.

Our follow-up question for North East London Cancer Alliance

Have faster X-ray reports at Barts, BHRUT and Homerton led to faster CT scans and diagnoses, and when will the evaluation be published?

Later developments (added Oct. 2026)

On March 23, 2026, RAD Magazine reported the same Barts figures. (RAD Magazine)
On March 24, 2026, UCLH reported results of LungIMPACT, which it called the largest randomised trial of its kind, published in Nature Medicine. Across five NHS trusts and 93,326 chest X-rays, using AI to flag abnormal scans for priority review did not significantly shorten the time to a lung cancer diagnosis: a median of 44 days with AI and 46 without, among 558 patients diagnosed. "We've shown that AI prioritisation, by itself, cannot fix that," said Nick Woznitza, the trial's principal investigator. The trial used qXR version 4.0 from Qure.ai, the same tool north east London uses, the Nature Medicine paper says, and ran from July 2023 to December 2024. Neither UCLH's release nor the paper identifies the north east London deployment as a trial site. (UCLH; Nature Medicine)
As of Oct. 25, 2026, we found no published results from the alliance's two-year evaluation.

Sources

North East London Cancer Alliance, “AI speeds up lung cancer detection in north east London,” Nov. 28, 2025. nelcanceralliance.nhs.uk
North East London Cancer Alliance, “Using Artificial Intelligence for faster chest X-ray results,” Nov. 4, 2024. nelcanceralliance.nhs.uk
North East London Cancer Alliance, funding announcement, Nov. 13, 2023. nelcanceralliance.nhs.uk
Barking, Havering and Redbridge University Hospitals NHS Trust, “AI helps reduce waits for chest x-ray results from weeks to days,” Nov. 5, 2024. bhrhospitals.nhs.uk
Qure.ai, “AI speeds up lung cancer detection in north east London.” qure.ai
RAD Magazine, “AI partnership helps speed up lung cancer detection across north east London,” March 23, 2026. radmagazine.com
University College London Hospitals NHS Foundation Trust, “AI-driven technology does not speed up lung cancer diagnosis,” March 24, 2026. uclh.nhs.uk
Nature Medicine, LungIMPACT trial, published online March 24, 2026. nature.com
HTN Health Tech News, “AI and imaging help tackle cancer waiting lists in Barking, Havering and Redbridge,” Nov. 27, 2024. htn.co.uk

Disclosure: [not in copy] Editor placeholder