Artificial intelligence can accurately identify breast cancer

By Tom Campbell via SWNS

Artificial intelligence could help screen for breast cancer - but doctors are still needed to catch cases that slip through the net, according to a new study.

AI identified more than 85 percent of cancers in the study. In the future, it could be used to save lives and significantly reduce radiologists' workload.

Women between the ages of 50 and 70 are encouraged to have their breasts screened for cancer frequently.

This creates a lot of work for radiologists who must carefully review millions of x-rays, known as mammograms, every year.

These scans are often independently checked to make sure no signs of cancer were missed the first time.

Now researchers have come up with a solution that could save radiologists having to second guess themselves.

Author Dr. Solveig Hofvind at the Cancer Registry of Norway in Oslo said: "In our study, we assumed that all cancer cases selected by the AI system were detected.

"This might not be true in a real screening setting.

"However, given that assumption, AI will probably be of great value in interpretation of screening mammograms in the future.”

A commercially available AI system was used to analyze 123,000 examinations performed on more than 47,000 women at four BreastScreen facilities in Norway.

In total, 752 cancers had been detected during the examinations, while 205 were discovered between screenings.

The AI system predicted people's risk of cancer on a scale of one to ten, with one representing the lowest risk.

It returned a score of 10 for 87.6 percent of the cancers which were spotted during screenings and 44.9 percent of those found during intervals.

Three thresholds were then created to assess the performance of the AI system as a decision-making tool.

When using the thresholds, the number of cancers which went undetected by the algorithm fell below 20 percent.

The AI tool could "substantially reduce" the number of mammograms to be reviewed, while only letting a small number of cancers slip through the net.

But these could then be picked up as part of a second screening performed by medical staff.

Dr. Hofvind said: "Based on our results, we expect AI to be of great value in the interpretation of screening mammograms in the future.

"We expect the greatest potential to be in reducing the reading volume by selecting negative examinations.”

More research is needed before AI can be rolled out to help with breast cancer screenings.

Dr. Hofvind added: "We are looking forward to testing out different scenarios for AI using retrospective data and then running a prospective trial."

The findings were published in the journal Radiology.