ELECTROCARDIOGRAM DIAGNOSIS PERFORMED BY ARTIFICIAL INTELLIGENCE

Abstract

Introduction: Cardiovascular diseases (CVD) have a significant impact on global health, requiring precise diagnosis and effective treatments. Electrocardiogram (ECG) is a fundamental tool, but its interpretation can be flawed. Artificial intelligence (AI) emerges as a promising alternative, offering greater accuracy and reliability. This integrative review explores the use of AI in ECG interpretation, aiming to enhance the therapeutic approach to CVD. Methods: This is an integrative review aimed at examining the application of AI in ECG interpretation within the context of CVD, conducted on the PubMed database with descriptors registered in DeCS, combined as specified, including free articles, Clinical Trials, Meta-Analyses, Randomized Studies, and Systematic Reviews, published in the last 5 years, in humans, in Portuguese, English, or Spanish languages, excluding duplicates and articles unrelated to the central theme. Results: Initially, 282 studies were identified, excluding 115 duplicate studies and another 161 based on title and abstract. For selection, 6 articles were included. Conclusion: The analysis revealed that AI has the potential to complement medical interpretation of ECG, offering effective diagnostic support. However, the effective implementation of AI faces significant challenges, including practical applicability issues, algorithm specificities, and ethical, legal, and social concerns (ELSI). It is essential to address these concerns to ensure ethical and responsible use of AI in clinical practice, despite the continuous need for robust studies and algorithm development to further strengthen its role in improving CVD diagnosis through ECG interpretation.