NHS trials AI technology to offer same-day diagnosis of aggressive lung cancer, which kills 35,000 Britons every year
- The NHS is testing AI technology to significantly speed up lung cancer diagnosis
- Experts believe that spotting lung cancer “days earlier” can help save patients’ lives
- It can take up to a week for experts to determine whether an X-ray shows fatal problems
- The software prioritizes which X-ray images the doctor should look at first
Artificial intelligence (AI) aims to reduce the time it takes NHS patients to receive a lung cancer diagnosis.
The program analyzes X-ray scans immediately after they are taken and alerts doctors if worrisome signs are noticed.
This means that the diagnosis can be reduced from a week to the same day.
Experts say early diagnosis is critical to treating lung cancer, which is so aggressive that catching it even a few days early can save lives.
Doctors examining X-rays for possible signs of lung cancer can take up to a week to make a diagnosis, which can harm a patient’s long-term prospects.
Developed by British Covid vaccine firm AstraZeneca and Indian tech firm Qure AI, the software is being rolled out across Manchester as part of a pilot study involving more than 250,000 people over the next six months.
Developed by British Covid vaccine firm AstraZeneca and Indian tech firm Qure AI, the software is being rolled out in Manchester as part of a pilot study involving more than 250,000 people over the next six months. The doctors involved say that if successful, the system could be used across the NHS.
AstraZeneca is now the second maker of Covid vaccines to turn its attention to artificial intelligence, as experts predict all NHS doctors will soon be working with high-tech software.
Almost 50,000 Britons are diagnosed with lung cancer each year. Over 35,000 people die from the disease during this time, making it the biggest cause of cancer in the UK.
An X-ray is usually the first test done to look for signs of disease, which include growths in the lungs that could be tumors and fluid or swollen lymph nodes in the chest.
“I would say it’s the most difficult X-ray to analyse,” says Professor Matt Evison, consultant chest physician at the Greater Manchester Cancer Alliance.
“Because there’s so much going on in that part of the body — the spine and the lungs — you have to look very carefully. This means it can take up to a week for patients to receive their results.
“Given how aggressive lung cancer is, this wait can make a big difference in how effective the treatment will be.”
According to one study in the medical journal The Lancet, late diagnosis and treatment of lung cancer in England is responsible for nearly five percent of deaths from the disease.
AI software flags X-rays that show signs of disease, so doctors can take a closer look first.
“We still look at all scans,” says Professor Evison, “but this new system means we prioritize the most worrying ones. You can get the results the same day.
“This is more important than ever, given the enormous pressure on the health service at the moment. We need to make these diagnoses quickly, but we only have limited resources.
“If it works, I could see it being used across the country.”
The news comes just days after Covid vaccine developer BioNTech, which teamed up with pharmaceutical giant Pfizer to create the first approved US cut, said it had bought British artificial intelligence company InstaDeep for £562m. BioNTech plans to use the company’s technology to speed up the drug and vaccine development process.
Experts say AI will play an increasingly important role in the NHS over the next decade.
“Every doctor in the NHS will be trained to work with artificial intelligence in the near future,” says Professor Nick Howes, an AI expert at the University of Oxford. “This could be used to decide which treatment is most appropriate for a newly diagnosed cancer patient, or to create a much more efficient 111 telephone referral system.
“Artificial intelligence will never replace doctors, but it will be an auxiliary tool that speeds up the medical process.”