THE march of AI now seems unstoppable in every part of our lives, from choosing the TV we watch to making our trains run on time.
Cancer is no exception, with the high-powered technology already turbo-charging research, diagnosis and treatment of the disease.
AI is turbo-charging research, diagnosis and treatment for cancer[/caption]
Only last month, the NHS announced that pioneering artificial intelligence tools will be used on an “unprecedented scale” to help millions of patients receive “faster and smarter care”.
This includes identifying possible signs of breast cancer in a trial of 700,000 women. But the potential doesn’t stop there.
In August, scientists found that AI may soon detect throat cancer from the sound of your voice.
And last week, scientists revealed an AI-modelled personalised vaccine against melanoma skin cancer is one step closer to reality.
From new drugs to “robo docs”, this is how AI is changing the cancer world . . .
FOR SCIENTISTS . . .
PREVENTION and early detection are the most powerful weapons we have against cancer.
Research labs analyse tons of data every day, looking for patterns in genetic codes and compiling decades’ worth of information.
This can be laborious, so AI is a game-changer; it’s bread and butter stuff for supercomputers.
Understanding how cancer starts and how to kill tumours can help us treat mutations and prevent them in future.
Professor Nuria Lopez-Bigas, from Spain’s Institute for Research in Biomedicine, tells Sun on Sunday Health: “Before AI, we would look at a patient’s tumour and guess which mutations to experiment on and see what happened, but it was slow.
“There are sometimes thousands of mutations, most of them not causing the cancer.
“With AI we can analyse thousands of tumours from people around the world and identify patterns that indicate which mutations are the drivers.
“We are working years or even decades faster.”
King’s College London this year used ChatGPT-4 to discover new drug combinations to test against breast cancer.
It harvested information from old studies to suggest 12 untried drug combinations, seven of which appeared to work when tested on cancer cells in a lab.
The Institute for Cancer Research also trained AI to identify how different medicines change the physical shape of human cancer cells, with 99 per cent accuracy.
The team believes that it could slash the time it takes to do pre-human testing of medicines from three years to just three months.
FOR DOCTORS . . .
AI is now being used in hospitals to check moles[/caption]
Hospitals are also trialling how machine learning can help spot prostate cancers earlier[/caption]
AI’s potential for speeding up cancer diagnosis is already being embraced by the NHS.
It can “see” pixels many times smaller than a human hair, revealing worries on scans or test results that might be invisible to the naked eye or missed by busy doctors.
NHS watchdog Nice, the National Institute for Health and Care Excellence, this year encouraged hospitals to use AI for checking moles.
A detection system, called DERM, compares photos taken with a smartphone to a huge database of similar images to help doctors decide if moles are cancerous.
Meanwhile, hospitals are also trialling how machine learning can help spot lung, breast and prostate cancers earlier.
Key to its usefulness is that diagnosing the disease is about comparing old data showing what cancer looks like, to new data from the current patient.
AI can check fresh information against millions of previous results in seconds.
Dr Joe Barnett, a radiologist at the Royal Free NHS Trust in London, says: “It’s vital to a new lung cancer-screening programme offering CT scans to high-risk Brits. Screening can catch cases early and decrease patients’ risk of dying, but it requires a huge number of scans and we don’t have enough radiologists.
“AI is helping massively. Instead of needing two radiologists, we can use one plus the AI system.
“It can detect potential masses quickly, helping me to spend less time looking for lumps and freeing me up to think about if they might be cancerous.”
A two-year study involving 3,000 men will test similar technology for prostate cancer after trials showed it could detect smaller MRI abnormalities than humans alone.
Oregon Health and Science University has trained an algorithm to spot early signs of throat cancer by listening to changes in recordings of patients’ voices.
King’s College London this year used ChatGPT-4 to discover new drug combinations to test against breast cancer[/caption]
FOR PATIENTS
COMPUTERS cannot yet perform cancer treatments, but doctors say they will help them to make faster and better-informed decisions.
“Precision”, “personalised” and “tailored” are the buzzwords of the new era, which focuses on treating patients’ individual cancers based on their body and genetics.
This is great for reducing side-effects and boosting the chances of remission, but means a fast-growing list of choices for doctors.
The American Society of Clinical Oncology this year launched a ChatGPT-style tool to guide its members.
Doctors enter their patient’s details and test results and the bot gives them a shortlist of treatment options based on the latest science.
Dr Julie Gralow, breast cancer doctor and President of the Association for Clinical Oncology, says: “If you’re a busy doctor and your patient’s scans unexpectedly show signs of progression you’ve got to scramble to decide what to do next.
“AI won’t replace oncologists, but those who use it will replace those who don’t.”
Smart tech can also predict how potential new therapies will work by simulating the effects on “virtual patients”.
An Oxford University study last year trained AI to safely adjust medication doses, and similar technology can track how well patients are responding to their current treatment and suggest how to improve it in real-time.
Machine learning may even be incorporated into surgical robots, which are being used more often for cancer operations.
These could speed up decision-making and automate tasks like tying knots in stitches.
Leading robot maker Intuitiv is already using AI to record operations and feed back to surgeons which bits they could have done better.
And in July, scientists at Johns Hopkins University in America programmed a robot to perform the first ever imitation surgery without human help.
AI will also help to monitor patients as they recover, by linking wearable sensors tracking blood pressure and heart rates to doctors’ computers.
THE BEST IS YET TO COME
AMERICA’S Cancer Research Institute describes the AI and cancer match-up as an “emerging revolution” that will “help to reshape how we understand, diagnose, and treat it in ways never before thought possible”.
Prof Lopez-Bigas adds: “I have a feeling that the biggest leap is yet to come.”
Cancer screenings in England
CATCHING cancer early gives you the best chance of survival, and a huge part of that is attending regular screenings.
NHS programmes can help diagnose the disease, or your risk of it, and improve the likelihood of successful treatment.
There are three national screening programmes in England: cervical screening, breast screening and bowel screening.
“If you are eligible, please make every effort to have your screening test as they can detect a problem early, before you have any symptoms. ,” the NHS says.
“Finding out about a problem early can mean that treatment is more effective.”
Cervical screening
This is offered in England to people with a cervix aged 25 to 64 and is routinely carried out every three years up to the age of 49, and every five from 50 to 64.
Depending on the result, people may be recalled earlier.
During a cervical screening, samples are tested for high risk Human Papillomavirus (HPV), which causes nearly all cervical cancers.
Those that test positive are then analysed further.
Breast screening
Breast screenings, which involve an X-ray test called a mammogram that can spot cancers when they are too small to see or feel, are usually offered to women aged 50 to 71 in England.
But the NHS is trialling them for women under 50 if they have a high risk of developing breast cancer.
Bowel screening
This test detects whether patients are showing any early signs of cancer.
It is available to everyone aged 50 to 74, with at-home kits automatically sent out every two years, so make sure your GP has your correct address.
The test involves providing a small poo sample to be checked for tiny amounts of blood, which could be caused by cancer.
If you’re 75 or over, you can ask for a kit every two years by phoning the free bowel cancer screening helpline on 0800 707 60 60.
Source: NHS
