Artificial algorithms that predict atrial fibrillation using artificial intelligence, created at AUTH

The use of personalized techniques which are based on genetics, but also on IT has entered medical everyday life. Artificial intelligence and large databases have enabled the creation of “intelligent” algorithms that can predict the prevalence of diseases in the general population, help diagnose diseases and recommend the most appropriate treatment. All this is also applicable in the field of cardiology and already with the collaboration of the 3rd Cardiology Clinic of AUTH and the Department of Medical Informatics of AUTH such algorithms have been created that use artificial intelligence to predict atrial fibrillation, which is the most frequent arrhythmia in the western world.

The director of the 3rd Cardiology Clinic of AUTH points out the above, Vassilis Vassilikos speaking to APE-MBE on the occasion of the organization of the conference entitled “Arrhythmias Update 2022” whose work will be carried out September 2-3 in Thessaloniki.

“Digital medicine is now an everyday thing these days. Now artificial intelligence is starting to enter clinical practice as well. There are many models that use some indicators with the aim of predicting or prognosing a disease or diagnosing a disease. We have been involved in the past and now, and now all our work begins and finds application. We have had a long-standing collaboration with the Medical Informatics department at our school and have built such algorithms that use artificial intelligence to predict atrial fibrillation. Also in our age with big data, this information is valuable and everyone is aiming for this thing. And of course it has to do with how you use this information to draw your own conclusions or test your algorithms. We have developed some such techniques and in addition we are trying to build such databases” notes Mr. Vassilikos

Precision medicine

He also notes that precision medicine is gaining ground these days.In previous years, for example, we took 1000 people and said that a percentage of the smokers among those 1000 will have a heart attack. But there are smokers who don’t get heart attacks and there are non-smokers who do. Now there is a more personalized approach based precisely on things that we did not know about in previous years. There are new techniques, among which some also use gene analysis or artificial intelligence or other mapping techniques, imaging methods, etc. with which a tailor-made prediction is finally made for each patient. For example, we say that so-and-so has these characteristics, the device gave us this prediction, the other imaging gives us that prediction and therefore has a higher risk of having this, so we start to become more targeted in our treatments. So a different approach is now starting, but it will take a long time to change that.” says Mr. Vasilikos

Atrial fibrillation and Big Data

During the conference, as mentioned by Mr. Vasilikos, the issues related to the prediction of atrial fibrillation will be discussed, as well as how sudden death can be predicted using large databases.

“Data will be presented from the large databases of the US Army where they take various measurable parameters from the simple electrocardiogram, which are analyzed with special techniques. Then they enter some mathematical models that are used to predict some diseases, such as atrial fibrillation, which is an atrial arrhythmia in the electrocardiogram, a part of which, for example, the “P wave” shows the electrical stimulation of the atria, where it is produced the vaginal stimulus. Within this wave with special analyzes which are not visible to the naked eye, nor also with other types of approaches but with a completely different analysis philosophy, they find the existing “hidden” information. This information can now be found with the computer and with these results and correlate it with other things and make conclusions as to whether or not someone may have atrial fibrillation depending on what they have. There are corresponding predictive algorithms for sudden death. Of course, everyone works with different degrees of success. Not everything is perfect, but this is the future and we have the pleasure of having such people who have invented these things have worked for many years and can share with us their own data” adds Mr. Vassilikos.

Referring to the subject of sudden death, he notes that in our country, sudden death, especially among young people, has the same distribution as it has in other Western countries.

“We now know what the most common cause of sudden death is and we need to target it. The most common cause is related to age. At a young age, up to 20-25 years of age, the most common cause is hypertrophic cardiomyopathy, which is also the case in the rest of the world, as the ages increase, it also starts to be coronary heart disease. Young people who have hypertrophic cardiomyopathy don’t know it. That’s why there is a specific way to prevent it in athletes or young people who exercise, and that is the cardiogram,” adds Mr. Vassilikos.

RES-EMP

Read the News today and get the latest news.
Follow Skai.gr on Google News and be the first to know all the news.