Radiographers who are working through large caseloads usually take seven to 14 days to review a patient’s chest X-ray and find out whether there are any abnormalities, such as a cancerous tumour.

But in July they began trialling an AI algorithm, known as red dot chest X-ray, which completes this process in just seven seconds so patients can get their results back on the same day and begin treatment sooner if they need it.

Nikhil Bhuskute, Deputy Medical Director at Calderdale and Huddersfield NHS Foundation Trust, said it relieves pressure on staff who conduct around 65,000 chest X-rays a year and “significantly shortens” waiting times for patients.

Huddersfield Royal Infirmary staff show Shadow Technology Secretary Peter Kyle how they use an artificial intelligence programme to detect lung cancer in chest x-rays

“Around 95 per cent of patients don’t have anything show up on an X-ray,” he said. “They don’t have to live in that anxiety and fear about having cancer for 14 days. They can have an answer in seven seconds.”

He also said that within the next four months patients could be sent for a same-day CT scan if the algorithm detects an abnormality. That would lead to a quicker diagnosis for many lung cancer patients and improve their chance of survival.

“It also saves a large amount of money in the long run,” said Dr Bhuskute. “The quicker we diagnose them, the quicker they are treated and that requires less resources.”

Dr Bhuskute, who is also a consultant radiographer, said some clinicians were anxious about the technology when it was introduced but now see it as an effective tool.

At Huddersfield Royal Infirmary the algorithm, developed by a company called Behold.ai, is currently only used to analyse scans of patients who were referred by GPs and the results are then checked by a trained radiographer.

“We are continuously auditing the performance of the AI algorithm with humans, to give us more confidence,” said Dr Bhuskute. “So far, it has not disappointed us.”

When it was trialled in Somerset in 2021, radiologists agreed with the results of almost 98 per cent of scans when the algorithm gave patients the all clear. It also cut the average amount of time a patient waited for a follow-up scan, from seven days to 2.8 days.

Ruben Rasalingham, Chief Operating Officer at Behold.ai, said his team of developers are always working to improve the accuracy of the red dot chest X-ray platform.

“It isn’t a self-improving algorithm like chat GPT. It’s a controlled medical device that is regulated,” he said. “We’re registered with the Care Quality Commission and have to make sure we’re always constantly improving the algorithm.”

He also said all of the medical data is anonymised before his company processes it at a centre in Reading.

It comes as hospitals across the country are preparing for a surge in demand over the winter months and attempting to tackle large waiting lists.

A record 7.75m people in England are waiting to start routine hospital treatment – the highest number since records began in 2007.

NHS staff at Yorkshire hospital confident AI technology can cut waiting lists and save lives