Advertisment

Google's AI Outperforms Radiologists At Detecting Breast Cancer

A new artificial intelligence program may outperform radiologists at detecting breast cancer in mammograms. The software, developed by Google Health, has the potential to improve accuracy in detecting breast cancer.

author-image
Prapti Sarkar
New Update
breast cancer awarenes, five-minute breast cancer treatment

A new artificial intelligence program may outperform radiologists at detecting breast cancer in mammograms. The software, developed by Google Health, has the potential to improve accuracy in detecting breast cancer.

Advertisment

How The AI Detected Breast Cancer

A team of researchers from Google Health and Imperial College London designed and trained a computer model. They used X-ray images from nearly 29,000 women. In a paper published in the scientific journal, Nature, the researchers reported that the AI is as good as radiologists in detecting breast cancer. The software also promises to reduce errors.  Further, in the paper, Google claimed to provide “evidence of the ability of the system to generalise from the UK to the US.”

The study was developed with Alphabet Inc’s DeepMind AI unit, which merged with Google Health in September last year. The findings represent a major leap in the potential for the early detection of breast cancer. The team trained the AI to scan X-ray images of mammograms. Then the AI looked for signs of breast cancer by identifying changes in the mammograms of the 29,000 women. The researchers then tallied the computer’s guesses against the women's’ actual medical outcomes.

Also Read: In India, Over 87,090 Women Die Due To Breast Cancer 

“In an independent study of six radiologists, the AI system outperformed all of the human readers: the area under the receiver operating characteristic curve (AUC-ROC) for the AI system was greater than the AUC-ROC for the average radiologist by an absolute margin of 11.5%," Google wrote in a blog post. According to Christopher Kelly, a scientist at Google, “the model performs better than an individual radiologist in both the UK and the US”.

However, while the AI outperformed doctors in most cases, it was still not perfect. There were many instances where doctors flagged breast cancer and the AI missed.

Advertisment

Can It Replace Radiologists?

However, while the AI outperformed doctors in most cases, it was still not perfect. There were many instances where doctors flagged breast cancer and the AI missed. “Sometimes, all six U.S. readers caught a cancer that slipped past the AI, and vice versa,” said Mozziyar Etemadi, a co-author of the paper. Thus, Google says that the software is one that can help but not replace radiologists.

According to Shravya Shetty, a Google researcher, both software and doctors bring their own strength. She believes that the process can be “complementary”.

“There are a number of cases where the radiologists catch something that the model misses, and vice versa. Bringing the two together could strengthen the overall results,” Shetty said.

Sara Hiom, director of cancer intelligence and early diagnosis at CRUK, told the BBC: "This is promising early research which suggests that in future it may be possible to make screening more accurate and efficient, which means less waiting and worrying for patients, and better outcomes."

“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.

Advertisment

What’s Next?

However, there is still a lot to do before the software can become a part of the process. The next step, according to Chris Kelly, a clinician scientist at Google Health, is to assess the AI in real-world conditions. Its performance could slip when it is fed images from different mammogram systems. In this particular study, nearly all of the images came from machines provided by one manufacturer.

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.”

The AI may not replace radiologists but it can widely aid them. They will not replace the staffing crisis. However, they can act as a safety net and a second pair of eyes.

Also Read: What Do Breast Cancer Patients Want? 

Picture Credit:  Maurer foundation

Advertisment

Prapti is an intern at SheThePeople.TV

Google AI Google Health breast cancer artificial intelligence
Advertisment