A new study has shown that an artificial intelligence tool can accurately predict the risk of cardiovascular disease and human death in less than a minute.
The I.I. tool allows the use of a non-invasive screening method, which is not necessary at the clinic; it will be particularly useful for people with medium- and high-risk circulatory diseases.
"The tool provides risk-level information in 60 seconds or less," said The Guardian Professor Alicia Rudnicka, the lead author of the study. "If the patient learns that he is at risk, he may be prescribed statinas or offered other treatment."
Diseases of the circulatory system, such as cardiovascular diseases, ischaemic heart disease, heart failure and stroke, are leading causes of morbidity and mortality around the world. Although there are numerous risk assessment systems, they do not always accurately identify those who develop diseases of the circulatory system.
The designers have shown that the width of the veins and arteries in the retina can indicate early circulatory disease and is fairly accurate. With this in mind, engineers have developed a fully automated tool supported by artificial intelligence called Quartz. It is used to assess the potential for visualization of retina vessels and known risk factors to predict the health of the vessels.
As part of the experiment, scientists used Quartz to scan retina images of 88,052 people, all British biobank participants aged between 40 and 69, and then studied the retinal images of 7,411 European cancer researchers in Norfolk, all between 48 and 92.
The health of the test subjects was monitored for seven to nine years, and the results showed that the width, curvature and change in the width of the veins and arteries in the retina are important predictors of death from cardiovascular diseases in men.
In addition, researchers found a strong correlation between the retina data calculated by Quartz and cardiovascular diseases, their deaths and strokes.
According to the authors of the study, prognosis of the risk of vasculometrics using I.E. is completely automated, inexpensive and non-invasive.