They develop a method to detect cancer 5 years before it appears

Despite the scientific advances of this century, in the vast majority of patients with breast cancer, the diagnosis is a surprise. For some, it's too late. Therefore, the identification of patients at risk before the disease develops has been the goal of researchers and effective early detection programs.

(You may also like: They discover that coffee can inhibit a type of cancer)


A team from the Artificial Intelligence Laboratory (CSAIL) and Massachusetts General Hospital (MGH) at MIT has created a new deep learning model that can predict from a mammogram if a patient is likely to develop breast cancer in the future.

The model was fed with the results obtained from more than 60,000 patients treated at MGH. Thus, he learned the subtle patterns in breast tissue that are precursors of malignancy.

Different positions

While it has been shown that mammography reduces mortality from breast cancer, There is a continuing debate about when to start the screening and how often to do it.

While the American Cancer Society recommend an annual evaluation from 45 years, the US Preventive Task Force UU Recommends ua biannual evaluation after 50 years.


In an article published by MIT, the professor of the Regina Barzilay Institute, breast cancer survivorHe explained that they expect these systems allow doctors to customize screening and prevention programs at the individual level, making late diagnosis something of the past.

What do you think of this discovery? Leave us your comments.