The use of Artificial Intelligence in heart disease: from prevention to treatment

• DOI: https://doi.org/10.22533/at.ed.1595352524124

• Palavras-chave: Artificial intelligence, heart disease, exams.

• Keywords: Artificial intelligence, heart disease, exams.

• Abstract:

  The concept of Artificial Intelligence (AI) was developed by Alan Turing in 1950 and consists of the idea of a digital mind capable of learning, adapting, reacting, and "thinking" in the same way as a human being, characterizing a concept called machine learning (ML) (OZSAHIN et al., 2022). This entire process is carried out through supervised and unsupervised learning methods of the AI computational model (OZSAHIN et al., 2022). In this way, this model receives data and "learns" how to create associations and differences between the information that has been entered, and as a result, performs tasks that classical computer programming is not capable of (OZSAHIN et al., 2022).
  The capacity for AI to act within the clinical and medical sphere shows promising effects in complementing human reasoning (TOLU-AKINNAWO et al., 2025). AI can act as an additional tool that assists doctors' judgment, i.e., it helps improve the quality of medical care, being responsible for making early and non-invasive diagnoses in as many patients as possible, increasing the quality of health care, reducing hospital costs, and making medical assistance more accessible to all (OZSAHIN et al., 2022). Its application covers multiple aspects, such as cardiology, including diagnostic imaging methods such as chest X-rays, echocardiography, and cardiac computed tomography; genetic evaluation; risk stratification processes based on the analysis of health system data, taking into account clinical history and tests performed; electrocardiographic records; and evaluation of cardiac auscultatory records from databases (OZSAHIN et al., 2022). 
  Thus, the justification for its use is relevant in various contexts, such as heart disease, given the prevalence and mortality of these diseases today (OZSAHIN et al., 2022). This is shown by the Centers for Disease Control and Prevention (CDC), which found that 20% of deaths in the US were caused by heart disease in 2020, equivalent to 696,962 people and characterizing the leading cause of death in the country (OZSAHIN et al., 2022). In addition, population growth and increased life expectancy further increase the workload in health centers, which, as a result, increases the number of medical errors that could be preventable (OZSAHIN et al., 2022).

  Therefore, in many cases, the integration of artificial intelligence and clinical cardiology practice can contribute significantly to improving cardiology care (GANDHI et al., 2018). Based on this analysis, it is important to note that even though rapid diagnosis can improve clinical decision-making and prevent serious cardiac complications, there are still many challenges, such as algorithm transparency and the need for continuous validation for their application (GANDHI et al., 2018). Therefore,  for AI algorithms to be widely implemented in clinical practice, they must have a result accuracy similar to or superior to that of human observers , reinforcing the importance of future studies that focus not only on developing new algorithms, but also on analyzing existing ones and verifying what can be done to implement them in the context of heart disease (OEVER et al., 2020).