Artificial Intelligence that Predicts Breast Cancer Risk

Breast cancer, which is generally seen in women, is a very serious disease and can lead to death like all cancer types.

A new artificial intelligence gives better results than all the models we have in this regard. Artificial intelligence, which examines the changes in breast tissue, evaluates the risk of developing breast cancer within five years.

8daa2738ca015737ac85fcfe902d589a05d5c2bf

More than 90,000 mammograms (scan results) were used to train the new deep learning model. The system noticed tiny patterns in these mammogram films that humans couldn’t see. These patterns show repetitive structures formed by structural deterioration. Artificial intelligence has discovered that 31% of people in the high-risk group are at risk of developing breast cancer in the future. This rate may seem low at first glance, but the success of doctors in diagnosing breast cancer at an early stage is around 18%.

Constance Lehman, one of the study’s authors, said that since the 1960s, radiologists have found breast tissues with various and distinctive patterns in women with breast cancer. Stating that these patterns may be the result of genetics, hormones, pregnancy, breastfeeding, diet and weight gain, the researcher said that they can now use this information for each woman separately and with a certain certainty.

Another advantage of the new artificial intelligence is that it can detect cancer risk in minorities with equal success. Previous methods were less successful in detecting breast cancer in people of color.

9e81417791b5c9aad3d55aed34cb10b6d1de3fc5

Artificial intelligence-based diagnostic methods have been developing in recent years. From Alzheimer’s to child depression, it is possible to diagnose using artificial intelligence in many areas. Researchers are now developing a new AI model for the system to gain widespread acceptance. It gives great hope for artificial intelligence to see patterns that the human eye cannot see. Artificial intelligence will evolve as we start using larger data sets.

The new research has been published in the journal Radiology.

Related Posts

Leave a Reply

Your email address will not be published.