AI SAY

‘Landmark’ moment as NHS clinics use AI to ‘detect breast cancer cases earlier and faster’

Tech will compare breast scans to past ones in an instant

NHS clinics will use artificial intelligence to screen for breast cancer in a massive “landmark” trial.

Almost 700,000 British women are expected to take part in a study of using algorithms to spot tumours earlier and faster.

Advertisement
Artificial intelligence can learn how to spot cancer by comparing to past scans (stock image)

The technology can check a breast scan against a database of thousands of previous scans to look for abnormalities that might signal cancer.

Health chiefs reckon it could allow one radiologist to do the work of two and spot more cancers in the early stages.

The rollout will start with an £11million trial in 30 breast clinics in England and extend to the entire NHS if it succeeds.

Professor Lucy Chappell, chief scientific adviser to the Department of Health, said: “This landmark trial could lead to a significant step forward in the early detection of breast cancer, offering women faster, more accurate diagnoses when it matters most.

Advertisement

“It is another example of how public-funded UK research is crucial for rigorously testing world-leading new technologies such as AI that can potentially save lives while reducing the burden on the NHS.”

AI could revolutionise healthcare, speeding up diagnosis and freeing up doctors' time

Dr Katharine HallidayThe Royal College of Radiologists

Breast cancer is the most common form of the disease in women with 57,000 new cases per year.

Three quarters of patients live for 10 years or more after diagnosis but catching tumours early is the key to boosting survival.

All women aged between 50 and 71, or younger if they are at high risk, get free routine screening on the NHS.

Advertisement

Most read in Health

FAT JAB BLITZ
Major change to online fat jab sales to stop people pretending to be obese
ASK DR ZOE
I suffer depression and love my friends but often feel worse after seeing them
'GAME-CHANGER'
New £20 DIY health test could prevent 136,000 deaths - it takes 10 minutes
‘WHY ME?'
I've had skin cancer twice and two brain tumours, I'm 18 - I fear I'll die young

It comes after the Royal College of Radiologists, who are responsible for reading scans, said Britain must urgently embrace AI in healthcare.

About 30 per cent of job posts are vacant meaning technology is vital to speed up testing.

Breast cancer symptoms you should NEVER ignore, with Dr Philippa Kaye

President Dr Katharine Halliday said: “AI could revolutionise healthcare, boosting productivity, speeding up diagnoses, and freeing up doctors’ time to focus on patients.”

Health Secretary Wes Streeting will also today commit to publishing a National Cancer Plan this year.

Advertisement

He will launch a call for evidence for the plan on World Cancer Day on Tuesday.

HOW COULD AI SPOT CANCER?

ARTIFICIAL intelligence is a big hope for the future of medical care and diagnosing cancer could be one of its most promising uses.

The amazing power of AI is in its ability to instantly remember and compare huge amounts of data and information that it has seen before.

For medical scans, an AI algorithm can be trained – meaning it is programmed to remember – with thousands of images taken from past scans.

If it is told during programming which scans were of patients with cancer and which were not, the algorithm learns to recognise cancer by linking certain patterns or results to a positive cancer diagnosis.

Then when it is shown a new image, it can detect whether cancer cells are present by comparing it to patterns of cells in every other scan image it has seen – like spot the difference.

Because the computer sees a picture as data, unlike an eye which sees light and dark, this also allows the AI to detect tiny changes that a human eye could not see – potentially allowing it to detect cancer earlier than a medical professional.

Use of the technology is still in its early stages and it must be closely supervised by real medics to make sure the algorithm is accurate.

Topics
Advertisement
machibet777.com