Health researchers at Houston Methodist Hospital have developed an artificial intelligence application they cite as being 30 times faster than human doctors and 99% accurate in reading and interpreting mammograms.

Houston Methodist used the artificial intelligence software to evaluate mammograms and pathology reports of 500 breast cancer patients. The software scanned patient charts, collected diagnostic features and correlated mammogram findings for breast cancer. Clinicians then used the results, like the expression of tumor proteins, to accurately predict each patients probability of a breast cancer diagnosis, the hospital says.

In the U.S., 12.1 million mammograms are performed annually, according to the Centers for Disease Control and Prevention. But 50% of mammograms yield false positive results, according to the American Cancer Society, resulting in one in every two healthy women tested being told they have cancer.

When mammograms fall into the suspicious category, doctors recommend biopsies for a broad range patients: those whose risk varies up to 95%. Of the 1.6 million breast biopsies performed annually nationwide, about 20% are unnecessary due to false-positive mammogram results, says the American Cancer Society.

Houston Methodist researchers say the artificial intelligence software they developed will help physicians better define the percent of risk that justifies a biopsy, eliminate unnecessary surgeries and obtain faster results. Manual review of 50 charts took two clinicians 50-70 hours, but artificial intelligence reviewed 500 charts in a few hours, saving over 500 physician hours, the hospital says.


This software intelligently reviews millions of records in a short amount of time, enabling us to determine breast cancer risk more efficiently using a patients mammogram, says chair of the department of systems medicine and bioengineering at Houston Methodist Research Institute Stephen Wong. This has the potential to decrease unnecessary biopsies.

The initial artificial intelligence results were released in August and now Houston Methodist Hospital researchers are looking at tens of thousands of images to see how the software can help with better detection of stage four breast cancerbreast cancer that has spread beyond the breast and nearby lymph nodes to other organs of the body, such as the lungs, distant lymph nodes, skin, bones, liver, or brain.

Houston Methodist Hospital is looking into obtaining a commercial license to potentially sell its artificial intelligence application for breast cancer detection, a process that will take about a year, says a hospital spokeswoman.