Modern possibilities of artificial intelligence in cardiology: from diagnosing pericardial diseases to personalized risk stratification in cardiomyopathy
TL;DR
Current data on AI in the diagnosis of myocardial and pericardial diseases and outcome prediction and key conditions for its implementation are algorithm explainability, data standardization, and prospective studies are summarized.
Abstract
Artificial intelligence (AI) is transforming cardiology, enabling early diagnosis and risk stratification in myocardial and pericardial diseases (dilated cardiomyopathy, myopericarditis, constrictive pericarditis). The purpose — to summarize current data on AI in the diagnosis of myocardial and pericardial diseases and outcome prediction. Material and methods. Literature search in PubMed, Scopus, Google Scholar (2021–2025) by relevant keywords, including systematic reviews and clinical guidelines. Results. AI algorithms are effective in ECG interpretation, image analysis and outcome prediction. In pericardial diseases, deep learning models achieve AUC >0.95 for differential diagnosis of constrictive pericarditis vs restrictive cardiomyopathy. In DCM, AI identifies phenotypes, predicts arrhythmic events and end-stage heart failure. Conclusion. AI is a promising tool in cardiology. Key conditions for its implementation are algorithm explainability, data standardization, and prospective studies.