The Use of Artificial Intelligence (AI) as a Cardiac Magnetic Resonance Image Analysis Engine to Detect Inflammation and Scarring in the Heart Muscle: A Systematic Review

Abstract

INTRODUCTION: Cardiac magnetic resonance imaging (CMRI) is a noninvasive imaging technique that allows a detailed assessment of the structure and function of the heart, being useful in identifying conditions such as myocarditis and cardiomyopathies. The application of Artificial Intelligence (AI) in the analysis of these images promises to increase diagnostic accuracy and efficiency in clinical management, as machine learning algorithms can identify subtle patterns that may go unnoticed by human observers. OBJECTIVE: The main objective of this review is to evaluate the effectiveness of Artificial Intelligence (AI) in the analysis of Cardiac Magnetic Resonance (CMRI) images. METHODOLOGY: This systematic review article, structured according to the PRISMA guidelines, aims to identify, evaluate and synthesize evidence on the use of Artificial Intelligence (AI) in the analysis of Cardiac Magnetic Resonance (CMRI) images to detect inflammation and scarring in the heart muscle. The review covered several steps, including the definition of the research question, study selection, data extraction and critical analysis of the results. Data analysis focused on identifying recurring patterns and themes, and the methodological quality of the studies was assessed using tools such as AMSTAR and the Newcastle-Ottawa Scale. RESULTS: The systematic review showed that AI algorithms, especially convolutional neural networks (CNNs), significantly improve diagnostic accuracy compared to conventional analysis. AI can achieve sensitivity and specificity rates greater than 90% in detecting myocardial inflammation and scarring. The implementation of AI in CMR analysis has the potential to transform the clinical management of patients with cardiac diseases, allowing for faster and more targeted interventions. CONCLUSION: The systematic review showed that the application of Artificial Intelligence (AI) in the analysis of cardiac magnetic resonance (CMR) images significantly improves the detection of myocardial inflammation and scarring. Machine learning algorithms, especially convolutional neural networks, demonstrated greater efficacy compared to conventional analysis by radiologists, allowing for faster and more effective interventions.