Artificial Intelligence Better At Predicting Breast Cancer than Doctors

“AI system could provide a second opinion without having the expense and difficulty of finding a second radiologist.”


A new study, led by Google Health, has found that an artificial intelligence (AI) system is much better at predicting breast cancer than radiologists.

Researchers behind this study opine that the technology could help improve cancer care and clinical outcomes.

Detecting cancer in its early stages could help in early treatment, improving the chances of survival, which is why most nations advise women to undergo routine screenings for any signs of breast cancer using a mammogram.

The American Cancer Society advises annual mammograms for women between the ages of 45 and 54. It also recommends women aged 55 and above to get screened at least once in a year or two.

In the United Kingdom, women between the ages of 50 and 71 are advised to get screened every three years.

However, such screening programmers are never perfect. And radiologists may miss same signs of cancer in some women.

In the United States, it is estimated that half of the women who have a mammogram will have a false positive result over a period of 10 years.

So in order to improve diagnosis and prognosis, Senior Staff Software Engineer at Google Health, Sharvya Shetty, and her team used an AI system on more than 90,000 mammograms from women in the US and UK. The participants were followed up for two to three years to confirm whether they developed breast cancer.

The researchers found that the AI system delivered 5.7 percent and 1.2 percent fewer false-positive results in the US samples and the UK samples, respectively, when compared with radiologists’ opinions.

They conducted another experiment, which showed that the AI system performed 11.5 percent better than the radiologists.

Adam Brentnall from the Queen Mary University of London said, “The system doesn’t do better than two radiologists working together – the number required to assess mammograms in the UK and other European countries. But the results appear promising.”

Dr. Dominic King at Google Health said, “The AI system could provide a second opinion without having the expense and difficulty of finding a second radiologist.” The Google Health team has also been investigating other uses of the AI system.