The researchers developed an algorithm that used a person’s image to predict biological life and survive in cancer patients. The results of technology, called a confrontation, on Thursday (8) in Lancet Digital Health.
From the point of view of researchers, the work shows that the image, like a simple personal image, contains important information that can help in clinical decisions and treatments for patients and doctors.
The biological age is a condition that reflects the functional state of the human body, and is not necessarily in line with a person’s temporal age. For example, a stable individual, who is a smoker and drinks alcohol, may have a more advanced biological age due to his unhealthy lifestyle of his time with regard to the history of birth.
According to the study, the face expectations that indicate that the most advanced biological age with the results of survival are worse in different types of cancer. They also found that algorithm exceeded doctors in short -term life expectations for patients who have received pouring radiation treatment.
“We can use artificial intelligence (AI) to estimate the person’s biological age from the face of the face, and our study shows that information can be clinically important,” said the study of senior study and correspondents of the study Hugo Aristus, director of the General AI program in a statement.
How was artificial intelligence developed?
Researchers at Mass General Brigham, US Hospital Research Company, used deep learning techniques and facial training. The tool has been trained with 58,851 photos of healthy individuals from general data groups.
The algorithm team in a group of 6,196 patients with two centers cancer, using routinely taken pictures at the beginning of radiotherapy.
The results showed that cancer patients seem to be much older than those who do not have cancer, and that, on average, their face was greater than their temporal age.
Among cancer patients, a more advanced face has been associated with worse survival results, especially in individuals who seemed more than 85 years old, even after adjustments to time, sex and type
The following steps
More research is needed before this technology is considered use in a real clinical environment. The team tests the algorithm to predict the disease, public health and the expectation.
Follow -up studies include -expanding this work to different hospitals, analyzing patients in different cancer stages, monitoring facial estimates over time and testing their accuracy in plastic surgery and makeup data groups.
“This opens the doors to a new field of discovering vital indicators of photographs, and its capabilities exceed the treatment of cancer or life prediction,” says a large study author for the study of Ray Mac, a member of the AIM program in Mass GENERAL BRIGHAM, in a statement.
“Since we are increasingly considered various chronic diseases such as aging diseases, it is extremely important to be precisely able to predict accurately with the path of aging for the individual. I hope that we can ultimately use this technology as an early discovery system in various applications, in a strong organizational structure and ethics, to help provide lives,” he adds.
IA in medicine: early diagnosis and risk reduction for patients