New AI tools can catch breast cancer early and predict which patients are more at risk of treatment side effects.
The duo of breakthroughs offer hope for the 55,000 British women diagnosed with the disease every year.
One of the tools, being tested in the UK, France and the Netherlands, opens up the door to giving patients more personalised treatment. It can calculate how likely a patient who has gone through surgery and radiotherapy will experience side effects, such as scarring, swelling of the arm (lymphoedema) and heart damage from radiation, years later.
Patients thought to be at higher risk could be offered alternative treatment, or extra support.
The tool – found to be around 73 per cent accurate at making three-year predictions – was trained on more than 6,000 breast cancer patients. The results were shared at the European Breast Cancer Conference in Milan.
Consultant breast surgeon Dr Tim Rattay, based at the University of Leicester, told The Guardian: ‘We hope this will assist doctors and patients in choosing options for radiation treatment and reduce side-effects for all patients.’
Dr Simon Vincent, director of research, support and influencing at Breast Cancer Now, said the tool was ‘exciting’.
He added: ‘This research is in its early stages and more evidence is needed before we can consider whether or not the AI tool could be used in medical settings, and we look forward to seeing results from the trial.’
Another AI tool called Mia successfully identified tiny signs of cancer in 11 women, which were missed by human doctors.
The tool was tested by NHS hospital Trust Grampian in Scotland. It analysed 10,000 mammograms.
When cancers are in their earliest stages they can be extremely small and hard to spot, with many invisible to the human eye. But many cancers can grow and spread rapidly.
One of the 11 patients whose cancer was flagged by AI tool Mia was Barbara.
Her cancer had been missed when studied by the hospital radiologists, the BBC reported.
Barbara’s 6mm tumour was caught so early by AI that she only needed five days of radiotherapy and an operation.
Because Barbara had not noticed any symptoms, her cancer might have not been spotted for another three years at her next mammogram appointment.
More than 90 per cent of women diagnosed with early-stage invasive breast cancer, with tumours smaller than 15mm, will survive the disease for five years or more, according to Cancer Research UK.
Tools like Mia could help to reduce waiting time for results, from two weeks to just three days.
The tool might even reduce workload as breast cancer doctors look at around 100 scans in a session.
During the trial each patient had a human review as well as one by Mia. However, it is hoped that instead of having two radiologists examine each scan, one radiologist could be replaced by this tool.
Dailymail.co.uk
