Andrea de VivarMadrid 05/08/2024 8:00 p.m.
Digital cardiology has its limitations.Pexels
The Artificial intelligence is a discipline in constant evolutionwhich seeks to improve the ability of machines to perform tasks autonomously, which has proven to be a useful tool in various areas of medicine and science. cardiology is no exception.
Today, advances allow the creation of algorithms and machine learning models that can analyze large amounts of data and help doctors in the diagnosis, treatment and prognosis of cardiovascular diseases.
Although Its use has many potential benefits, it also has limitations and some possible dangersaccording to a panel of experts who participated in the 2023 Congress of the European Society of Cardiology in Amsterdam.
Of its benefits…
Digital health and artificial intelligence would contribute to patients’ self-care. One of them is the teleconsultation, located among the most common uses of digital health, which is characterized by allow doctors to reach hard-to-reach placeswhether for reasons of distance or safety for an initial diagnosis.
Similarly, tele-interconsultation makes it possible to communication between health professionals that are located in distant cities or countries, which can be very useful and allows the creation of work teams called ‘teams’ to resolve complex cases, such as Heart Teams (where aortic valve implants or replacements are discussed, among other things,) or Neuro Vascular Teams (in patients with carotid disease or need for an intervention called ‘closure of patent foramen ovale’ in patients with stroke), are some of the examples.
{{ #cards }}
{{#section.link.href}} {{section.link.title}} {{/section.link.href}}
{{title.data}}
{{ /cards }}
Another instrument of digital health is devices called wearables in English, which generate data while the person is wearing themsuch as the ring known as Oura ring, which records how many hours we sleep and the duration of each of the stages of sleep to improve rest, or watches that detect the rhythm of heartbeats.
But when we talk about the imaging methods that we use daily in cardiology, the penetration of digital health is even greater: in magnetic resonanceFor example, there are tools that facilitate image acquisition with a significant reduction in time that the patient has to be inside the resonator (many patients cannot endure a long time inside it and this limits the study).
In cardiac tomography artificial intelligence allows us assess coronary flow non-invasivelysomething that before could only be evaluated through catheterization.
We cannot fail to mention telemonitoring platformswhich is the remote monitoring of patients with clinical conditions through a set of sensors associated with a digital platform that allows the necessary data on health status to be captured, processed and sent to a monitoring center where a team of medical professionals with extensive experience and specialization in each coverage area, performs monitoring. patient’s continuum; These platforms are currently applied to pathologies such as heart failure or cardiovascular rehabilitation.
…to its limitations
According to Alan Fraser of the University Hospital of Wales in Cardiff, The performance and accuracy of these devices varies greatly between them.. Fraser shared the stories of three patients whose devices no life-threatening cardiac events were detectedthus emphasizing that we still do not have evidence that wearable devices have a positive impact on patients.
ChatGPT has been shown to accurately answer simple cardiology questions, however, does not have the ability to respond to complex questions and reportscases in which it can be extremely unreliable.
There are, therefore, Big risks in using artificial intelligence to manage complex cardiology data, as it lacks the understanding of the clinical context necessary to interpret the data correctly. Furthermore, currently does not have the ability to differentiate between the need for acute versus chronic treatment and misses key concepts, such as essential laboratory tests and recommendations for contraindicated drugs.
