Artificial Intelligence Can Identify Breast Cancer Risk At 30 Times The Speed Of A Human

It has the potential to reduce biopsy numbers.

31/08/2016 12:26

Scientists are now using artificial intelligence to interpret mammograms and more accurately predict women’s risk of breast cancer.

The AI software, which has been developed by The Houston Methodist Research Institute, can translate mammogram charts into diagnostic information at 30 times the speed of a human. 


Doctors manually reviewing 50 mammogram charts took two people 50-70 working hours, but AI was able to review 500 in just a couple of hours.

Not only does it expedite the diagnoses, AI also has 99% accuracy in predicting the risk of breast cancer. 

In the USA, over 1.6 million biopsies are performed nationwide each year, and approximately 1 in 5 of these were unnecessary on cancer-free breasts.

Currently, when mammograms fall into the suspicious category, a broad range of 3 to 95 percent cancer risk, patients are recommended for biopsies.


The study, which was published in the Cancer journal, hopes to eliminate this 20% error margin and reduce the overall number of people going in for surgery.  

Stephen T. Wong at Houston Methodist Research Institute said: “This software intelligently reviews millions of records in a short amount of time, enabling us to determine breast cancer risk more efficiently using a patient’s mammogram. This has the potential to decrease unnecessary biopsies.”

Suggest a correction