New medical AI device can detect more breast cancer tissue – Deseret News

1 minute, 33 seconds Read

A trial in the U.K. has shown promise for a new medical AI device that can detect breast cancer that doctors couldn’t spot, according to a new press release from Microsoft. The trial also showed a decrease of workload and patient recalls.

Over 10,000 patients in the U.K.’s NHS Grampian health care network participated in the trial, where the medical AI device detected 12% more breast cancer tissue than doctors did alone.

The medical AI tool ‘Mia’

The medical trial used an AI tool known as Mia, short for Mammography Intelligent Assessment, created by the company Kheiron Medical Technologies. The BBC reports that Mia was used alongside NHS doctors as both analyzed over 10,000 mammograms.

Not only did Mia correctly identify all mammograms that showed symptoms of breast cancer, it also identified 11 additional mammograms with breast cancer that doctors did not identify, per BBC.

Business Insider explains that the cancer tissues Mia noticed were too small for doctors to notice, with one patient saying, “My cancer was so small that the doctors said it would not have been picked up by the human eye.”

Mia brings a faster diagnosis

The press statement explains that Mia was able to find 12% more breast cancer than the normal doctor analysis. With Mia, doctors were able to reduce the time it took to notify patients from 14 days to three days. Further, augmented AI workload revealed that Mia could help decrease the workload by 30%.

According to the National Cancer Institute, 20% of breast cancer is missed during mammogram screenings. Mia would help reduce those numbers by helping medical staff detect breast cancer at an earlier rate, per the press release.

According to Professor Lesley Anderson from the University of Aberdeen, “The tool would significantly improve breast screenings as more cancers would be detected without having to ask patients to take part in additional tests,” per Business Insider.

This post was originally published on this site

Similar Posts