The use of artificial intelligence for the diagnosis of diseases is increasingly common. Above all, algorithms capable of tracking X-ray images or MRIs in search of abnormalities indicative of a disorder are often used. He breast cancer It has been one of the most studied of its kind. But now artificial intelligence has been improved to the point of being able to predict its appearance even four years before.
At least, that’s what he explained to CNN Dr. Larry Norton, medical director at the Lauder Breast Center at Memorial Sloan Kettering Cancer Center. According to him, thanks to an artificial intelligence algorithm developed by the Massachusetts Institute of Technology (MIT), spots that could become breast cancer in the future can be detected.
This does not mean that artificial intelligence will replace radiologists. Not even that it can ensure 100% that these spots are going to transform into breast cancer. However, it can be a very useful tool so that radiologists decide what preventive measures patients can take well in advance.
Artificial intelligence to anticipate breast cancer
As with any other artificial intelligence algorithm, this one has been trained with mammography images of healthy patients who either developed breast cancer or never did. In this way, the machine is capable of capturing common factors in all those who became ill and detecting spots that, over time, they can turn into cancer.
As explained by Dr. Norton, it can be anticipated up to four years. But this does not eliminate the role of the radiologist. It will be this person who must later analyze the mammography and issue a diagnosis. If the possibility of cancer is finally confirmed, the patient can make decisions based on her risk.
If you also have mutated high-risk genes, you can take the option of performing a mastectomy. However, it may be sufficient to carry out regular tests such as MRIsto monitor the possible tumor.
In the interview, the medical director explains that, in this way, other more invasive tests can be avoidedsuch as biopsies, which are done regularly, can be very annoying.
AI won’t take jobs away from radiologists
Radiologists reviewing mammograms are known to miss one in eight breast cancer diagnoses. It is not something strange, because they do not stop being human, and sometimes it is very difficult to discern the tumor in the images.
For this reason, artificial intelligence is conceived simply as one more tool. Even so, these algorithms can also give false positives, so the last word will always be with the radiologist. So, the diagnosis will be improvedor, with the consequent benefit for the patients, who will be able to take action sooner. It is not about stealing work or replacing anyone, but about adding tools to the list that these professionals already have. Any help will always be welcome.