Heart attack, an algorithm can predict it. Thus artificial intelligence analyzes the risks of the heart

L’artificial intelligence, as is the case in other fields, is also acquiring considerable importance in cardiology. In particular with regard to some methods of increasingly frequent use such asultrasoundcomputed tomography (CT) or magnetic resonance imaging, the ability to acquire and reconstruct or process images represents a significant step forward in predicting the cardiovascular risk and perhaps also of the prognosis of our patients.

Heart attack and stroke, higher risk with cholesterol also for young people: how to reduce the risk of cardiovascular disease

An algorithm will also be able to predict heart attacks

In fact, among the various advantages of using artificial intelligence, in addition to speeding up processes, there is the reduction of human error and the variability between the various operators, increasing the accuracy of the diagnosis.

RESEARCH
In the latest issue of the European Heart Journal, Mary van Assen and her collaborators from the Radiology Department of Emory University in Atlanta (USA) have published an interesting article on the role of artificial intelligence in cardiovascular diagnostics. In echocardiography the evolution was evident. We have gone from devices that allowed us to evaluate the movement of the heart valves and walls to others that allow us to quickly analyze in depth the vitality of the various cardiac segments, flows and pressures, allowing us to optimize risk stratification.

SITUATIONS
As far as cardiac CT is concerned, many algorithms are being developed for a faster and more precise evaluation of the entire coronary tree. There are in fact situations (which are often present in diabetic patients or patients with renal insufficiency) in which the predictivity of coronary angioTac is significantly reduced. We are trying to remedy this both with new programs that limit the distortions of the signal due to calcium, and with algorithms that allow us to evaluate not only the anatomy of the coronary vessel, but also the flow of blood inside it. A kind of integration between two different methods that should significantly improve the diagnostic accuracy of the exam. In the future, it is very probable that algorithms will be developed which can make it possible to identify, through an angioTac, not only the extent of the coronary plaque, but also the danger, i.e. the probability that these plaques can break causing a myocardial infarction. Magnetic resonance imaging has only recently become part of the cardiologist’s diagnostic baggage. In fact, it can currently provide us with a large amount of data on cardiac function, valvular disease, muscle circulation and blood flow. Currently the resonance requires an expert operator to be able to treat the images that the machine provides us. However, machines that can carry out this delicate job are already in the test phase.

THE FUTURE
What do we expect from the future? That the various information obtained through different methods can increasingly be integrated into a single cardiac model. Personalized medicine is just around the corner.
In short, artificial intelligence is changing our vision of cardiology, playing an increasingly important role in the diagnostic process. However, the real future will be not to get too enthusiastic about the technique, but to exploit it intelligently for the good of our patients.

* Professor of Cardiology at the Catholic University of Rome