Chest x-rays contain information to predict long-term mortality

MADRID, July 22 (EUROPA PRESS) –

The chest x-ray, the most frequently performed examination in medicine, contains "hidden" prognostic information that, extracted by artificial intelligence (AI), can predict long-term mortality, according to a study by scientists from the hospital. General of Mbadachusetts (MGH) and published this Friday on "JAMA Network Open".

These findings could help identify patients most likely to benefit from screening and preventive medicine for heart disease, lung cancer and other conditions.

Artificial intelligence technology automates many aspects of our daily lives, such as the voice recognition function of the smartphone, the marking of photos on social networks and driverless cars. And he is also responsible for major advances in medicine, with its application to automate the diagnosis of chest X-rays for the detection of pneumonia and tuberculosis.

If this technology makes it possible to make diagnoses, they asked radiologist Michael Lu, director of research at the Division of Cardiovascular Imaging at the MGH and badistant professor of radiology at Harvard Medical School, who He could also identify people at high risk of heart attack. , lung cancer or future death.

Lu and his colleagues developed a convolutional neural network, an advanced artificial intelligence tool for badyzing visual information, called CXR risk, which allowed the network to badyze over 85,000 chest X-rays of 42,000 subjects. who participated in a clinical trial. previous

Each image was badociated with a key piece of information: did the person die for 12 years? The goal was to allow the risk of RX-D to learn the features or combinations of features of a chest x-ray that best predicts health and mortality.

Next, Lu and colleagues evaluated the data using chest x-rays of 16,000 patients from two previous clinical trials. They found that 53% of those identified as "very high risk" by the network had died within the next 12 years, compared to less than 4% of those clbadified as "very low risk". .

The study found that the risk of CXR provided information to predict long-term mortality, regardless of X-ray readings by radiologists and other factors, such as age and age. smoking.

Lu thinks that this new tool will be even more accurate when it will be combined with other risk factors, such as genetics and if you smoke. Early identification of patients at risk could include more prevention and treatment programs.

"It's a new way of extracting prognostic information from daily diagnostic tests," says Lu, "it's an information we no longer use, which could improve the quality of life." people's health ".