(Bloomberg) — Artificial intelligence can spot breast cancer more accurately than doctors, according to a study by Google Health.
The technology’s reading of mammograms reduced both false positives, where healthy patients are mistakenly diagnosed with the disease, and false negatives, where the cancer is missed, the Alphabet Inc. unit said in a blog post on Wednesday. The system reduced false positives by 5.7% in the U.S., according to data from more than 28,000 mammograms performed there and in the U.K.
Artificial intelligence is particularly good at reading scans, often outperforming experts. Last year, Google published research that showed how the technology could be used to tell whether breast cancer had spread to surrounding lymph nodes, helping pathologists make more accurate diagnoses.
Google is also training artificial intelligence to help determine whether a patient is likely to live or die, mining thousands of data points to help make predictions about outcomes.
Read more about Google’s medical research here.
Still, the company found it has to tread carefully when using patient data. In 2017, British regulators said Alphabet’s artificial intelligence unit, DeepMind, violated U.K. data-protection law when it tested an app that analyzed public medical records without telling patients how their information would be used.
The initial findings for the breast cancer study were published in the journal Nature. The research was done in partnership with DeepMind as well as Cancer Research U.K. Imperial Centre, Northwestern University and Royal Surrey County Hospital, the company said in the blog.
To contact the reporter on this story: Amy Thomson in London at [email protected]
To contact the editors responsible for this story: Giles Turner at [email protected], Marthe Fourcade, John Lauerman
<p class="canvas-atom canvas-text Mb(1.0em) Mb(0)–sm Mt(0.8em)–sm" type="text" content="For more articles like this, please visit us at bloomberg.com” data-reactid=”20″>For more articles like this, please visit us at bloomberg.com
©2020 Bloomberg L.P.
Add Comment