Comprehensive characterization of cardiac contraction for improved post-infarction risk assessment

Area of Science:

• Cardiology
• Medical Imaging
• Artificial Intelligence

Background:

• Assessing long-term outcomes after acute myocardial infarction (AMI) is crucial for patient management.
• Established Cardiac Magnetic Resonance (CMR) imaging provides valuable prognostic information but can be further enhanced.
• Novel methods for characterizing left ventricular contraction dynamics may offer improved risk stratification.

Purpose of the Study:

• To identify risk-related patterns in left ventricular contraction dynamics using novel volume transient characterization.
• To evaluate the prognostic value of three distinct CMR-based characterizations of cardiac function.
• To develop an enhanced CMR risk model for predicting major adverse cardiac events (MACE) in AMI survivors.

Main Methods:

• A multicenter cohort of 1021 AMI survivors underwent CMR imaging.
• Cardiac function was assessed via volume temporal transients, feature tracking strain analysis, and 3D shape analysis.
• A fully automated pipeline extracted conventional and AI-derived metrics, investigating their association with 12-month MACE.

Main Results:

• All three novel characterization methods demonstrated independent prognostic value beyond existing CMR indices, biomarkers, and risk factors.
• Combining the three approaches into a CMR risk model significantly improved MACE prediction by 13% (AUC 0.774 vs. 0.683).
• The automated pipeline successfully extracted metrics contributing to enhanced post-infarction risk assessment.

Conclusions:

• Novel characterization of left ventricular contraction dynamics provides significant prognostic information in AMI survivors.
• An automated CMR analysis pipeline enables advanced risk prediction for major adverse cardiac events.
• This approach enhances the utility of CMR in post-infarction patient assessment and management.