THE IMPACT OF ARTIFICIAL INTELLIGENCE (AI) INNOVATIONS IN CARDIOLOGY: A COMPREHENSIVE REVIEW OF CLINICAL EFFECTIVENESS, ETHICAL CHALLENGES, AND SOCIO-SYSTEMIC IMPLICATIONS

The aim of this review article is to provide an in-depth analysis of the current state of knowledge on the use of Artificial Intelligence (AI) in cardiology. We place particular emphasis on actual clinical effectiveness and the technological, ethical and social consequences of its implementation in healthcare systems. The review is based on carefully selected literature from 2018–2025, which was searched in the PubMed, Scopus and Web of Science databases, focusing on four key dimensions: technology, clinical practice, ethics and healthcare systems. AI (ML, DL) algorithms offer revolutionary potential in diagnostics (e.g., LVEF automation) and risk prediction (e.g., identification of hidden phenotypes in ECG). However, their implementation has often stalled due to critical non-clinical barriers. Among these, we identified: the problem of Explainability (XAI), which undermines legal accountability (especially with adaptive SaMD models); Algorithmic Bias, resulting from underrepresentation of training data, carrying the risk of exacerbating health inequalities (Health Equity); as well as serious systemic obstacles related to the need for continuous regulatory oversight (e.g., AI Act) and staff retraining. In order for us to reap the medical benefits of AI in a fair and safe manner, we urgently need policy recommendations on model validation across diverse cohorts, the development of transparent XAI architectures (rather than post-hoc methods), and the creation of flexible regulatory frameworks and educational programmes at the system level.