Medical Minute: AI Breast Cancer Screening

Artificial Intelligence seems to do a better job at predicting breast cancer than the standard clinical risk model, according to a new study in the journal Radiology.

Researchers from Kaiser Permanente Northern California applied five different A.I. algorithms to thousands of mammograms and found each one was better at predicting the five-year risk for breast cancer than the standard risk model known as "Breast Cancer Surveillance Consortium" (BCSC).

BCSC uses information including age, family history of the disease, whether the person has previously given birth and whether she has dense breasts to calculate a risk score. The A.I. algorithms use only the mammogram itself.

Among women with the highest risk, A.I. was able to predict up to 28%of cancers, compared to 21% predicted by the standard model.

Researchers add, combining the two risk models improved cancer prediction even more.

According to the Centers for Disease Control and Prevention, about 264,000 women and 2,400 men are diagnosed with breast cancer in the United States each year.