Artificial intelligence in the diagnosis and risk stratification of myocardial and pericardial diseases: two clinical observations.

Abstract

Pericarditis and dilated cardiomyopathy (DCM) are challenging conditions in clinical cardiology with high risk of life-threatening complications. Artificial intelligence (AI) technologies can substantially improve diagnostic accuracy and enable personalized risk assessment. Purpose: to evaluate, using two clinical cases, the hypothetical capabilities of AI for early diagnosis of myopericarditis and optimization of treatment strategy in a patient with DCM awaiting heart transplantation. Materials and methods. A retrospective analysis of two clinical cases from the cardiology department of SamGMU Clinics was performed. Decisions of a hypothetical AI clinical decision support system were modelled based on real clinical, laboratory, and instrumental data. Results. In the first case, AI analysis predicted myopericarditis with >85% probability based on the triad «effusion + troponin + inflammation» and identified a probable viral aetiology. In the second case, AI justified dual-chamber ICD implantation and personalized transplantation risk assessment: primary graft dysfunction 15–25%, high risk of infectious complications, and moderate risk of acute rejection. Findings. The clinician–AI synergy can accelerate diagnosis and individualize treatment in pericardial and myocardial diseases. Routine implementation requires addressing explainability, regulatory frameworks, and data standardization.