An artificial intelligence program developed by Google Health suggests AI is better at spotting breast cancer in mammograms than expert radiologists
This is an edited version of an article first published by The Guardian.
The AI outperformed the specialists by detecting cancers that the radiologists missed in the images, while ignoring features they falsely flagged as possible tumours.
If the program proves its worth in clinical trials, the software could make breast screening more effective and ease the burden on health services such as the NHS where radiologists are in short supply.
“This is a great demonstration of how these technologies can enable and augment the human expert,” said Dominic King, the UK lead at Google Health. “The AI system is saying ‘I think there may be an issue here, do you want to check?’”
About one in eight women are diagnosed with breast cancer in their lifetime. Screening programmes catch more than 18,000 cases each year in England alone, but tumours are still missed, giving false negative results. Some women are wrongly suspected of having cancer in false positives that lead to unnecessary anxiety and invasive biopsies.
Google’s AI program analyses mammograms in three different ways before combing the results to produce an overall risk score. The scientists trained the program on mammograms from more than 76,000 women in the UK and 15,000 women in the US. To see how well it worked, they then asked it to assess nearly 30,000 new mammograms from UK and US women who either had biopsy-confirmed cancer, or no signs of cancer during follow-up at least a year later.
In the US, women who go for breast cancer screening tend to be seen every one or two years and their mammograms are examined by a single radiologist. When compared with the US system, the AI produced 5.7% fewer false positives and 9.4% fewer false negatives.
In the UK, women are screened less often, typically once every three years, but their mammograms are reviewed by two radiologists, and sometimes a third in case of disagreement. The AI performed only marginally better than the UK system, reducing false positives by 1.2% and false negatives by 2.7%.
The results suggest the AI could boost the quality of breast cancer screening in the US and maintain the same level in the UK, with the AI assisting or replacing the second radiologist.
Breast cancer screening in the UK is under strain. The Royal College of Radiologists has identified a shortfall of at least 1,104 radiologists. In breast radiology specifically, 8% of hospital posts are unfilled, with much of the shortage due to older radiologists retiring from the NHS faster than new ones join.
Chris Kelly, a clinician scientist at Google Health, said the next major step would be a trial to assess the AI in real-world conditions. Its performance could slip when it is fed images from different mammogram systems. In the latest study, reported in Nature, nearly all of the images came from machines provided by one manufacturer.
“Like the rest of the health service, breast imaging, and UK radiology more widely, is understaffed and desperate for help,” said Dr Caroline Rubin, vice-president for clinical radiology at the Royal College of Radiologists. “AI programs will not solve the human staffing crisis, as radiologists and imaging teams do far more than just look at scans, but they will undoubtedly help by acting as a second pair of eyes and a safety net.
“It is a competitive market for developers and these programs will need to be rigorously tested and regulated first. The next step for promising products is for them to be used in clinical trials, evaluated in practice and used on patients screened in real-time, a process that will need to be overseen by the UK public health agencies that have overall responsibility for the breast screening programmes.”
Michelle Mitchell, Cancer Research UK’s chief executive, said: “Screening helps diagnose breast cancer at an early stage, when treatment is more likely to be successful, ensuring more people survive the disease. But it also has harms such as diagnosing cancers that would never have gone on to cause any problems and missing some cancers. This is still early stage research, but it shows how AI could improve breast cancer screening and ease pressure off the NHS.”