DeepMind (Google) developed AI to detect breast cancer

[ad_1]

After eye diseases, DeepMind and Google have developed a tool capable of detecting breast cancer, with more than encouraging results. What probably help specialist radiologists in the future.

DeepMind researchers have developed – in collaboration with Google Health teams – a tool based on artificial intelligence and machine learning able to identify breast cancer tumors on radiographic images. Subject of a publication in the scientific journal Nature, this AI would be as efficient as a human radiologist in this exercise.

She was trained using two databases (one British and the other American), containing a total of nearly 30,000 mammograms. In addition to fairly low false positive and false negative rates, this AI was able to be effective in its diagnostics in the face of six specialized radiologists, surpassing them on average by 11.5%.

A second automated reading

Mammograms can detect breast cancer at an early stage, but the results can sometimes be difficult to interpret and there are cases of false positives. For Michelle Mitchell, director of the British cancer research center interviewed by Business Insider, AI therefore has a role to play alongside doctors to improve detection, at a stage when treatments are more likely to succeed.

"These results highlight the important role that AI could play in the future of cancer care. The adoption of such technology could help improve the way we diagnose cancer in the years to come", she says before adding: "While more clinical studies are needed to see how and if this technology can work in practice, the first results are promising."

At DeepMind, we consider that such AI is not intended to replace radiologists, but that by providing a second automated reading, it can improve the detection rate of cancer cells. Google researchers will of course continue to develop and improve these types of tools. Recall that DeepMind had already looked into the creation of AI models to identify eye diseases.

[ad_2]