Google, Northwestern Medicine partner to develop AI tool to triage breast cancer patients

Listen
Text
  • Small
  • Medium
  • Large

Google and Chicago-based Northwestern Medicine are teaming up to develop an artificial intelligence tool to support clinician work by triaging mammography patients, according to a May 18 Google press conference.

One in 8 women will face a breast cancer diagnosis. Mammograms can help patients catch a diagnosis earlier, but half of all women have experienced a false alarm in the last decade, a Google spokesperson said. 

In addition to false alarms, it can take days to weeks for patients to get their results back from radiologists because of shortages in the field. 

"This is a national problem," Sarah Friedewald, MD, lead of the study and a physician at Northwestern Medicine, said. "We don't have enough people doing what we need to do."

The AI tool is expected to detect signs of breast cancer in mammograms at a similar accuracy as clinicians. It is also expected to alleviate the anxiety associated with patients having to experience long wait times for results.

The AI tool will analyze mammogram images and flag potential patients who may need more imaging to confirm a diagnosis. A clinician will review the flags and if the clinician agrees, they will contact the patient for more mammograms.

This system is supposed to help patients get through the mammogram and triage process much quicker, but is not expected to be released for several years.

"We are very much in the beginning stages of this research," Dr. Friedewald told Becker's. "It will be at least a year before we see the results of this study. In addition, there are likely several other studies that will need to be done after this one, before we can even submit for FDA approval. Therefore, it will likely be several years before we can entertain submitting to the FDA."

Copyright © 2021 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.

 

Featured Whitepapers

Featured Webinars