Abstract: The MBF + GLS model allowed more accurate risk stratification of STEMI after PCI than visual qualitative analysis. The DNN-assisted MCE quantitative analysis is an objective, efficient and reproducible method to evaluate microvascular perfusion.
Introduction
Predicting outcomes after a ST-segment elevation myocardial infarction (STEMI) remains a significant clinical challenge. While percutaneous coronary intervention (PCI) successfully restores flow in large epicardial arteries, microvascular obstruction (MVO) often persists, leading to major adverse cardiovascular events (MACE). Traditional Myocardial Contrast Echocardiography (MCE) relies on subjective visual assessment, which is prone to observer variability and often overestimates risk. This study introduces an AI-driven framework that combines automated microvascular blood flow (MBF) analysis with global longitudinal strain (GLS) to provide a more objective and accurate prognostic tool.
Methods
The researchers developed a multi-step pipeline to quantify heart health using a Deep Neural Network (DNN). The architecture utilizes a U-Net-based encoder with hierarchical Convolutional LSTMs to ensure temporal consistency across MCE frames. This system automatically segments the myocardium and calculates MBF by fitting time-intensity replenishment curves.
By integrating MBF with GLS—a measure of mechanical pump function—the team created a comprehensive risk prediction model. This approach captures both the "plumbing" (vascular perfusion) and the "pump" (mechanical contraction) aspects of cardiac recovery, moving beyond simple visual "eyeballing" of contrast flow.
Study Population
The DNN was initially trained and validated using an internal dataset of 7,403 MCE frames. The clinical utility was then assessed in a cohort of 194 STEMI patients who underwent successful PCI and were followed for at least six months to monitor for MACE.
Key Results
The combined MBF + GLS model demonstrated exceptional prognostic power, achieving an Area Under the Curve (AUC) of 0.95 for predicting MACE, significantly higher than the 0.70 achieved by traditional visual methods. At a 40% risk threshold, the model exhibited 84% sensitivity and 94% specificity. Patients identified as "high risk" by the model had a 91.7% event rate, compared to only 4.2% in the "low risk" group.
Beyond accuracy, the AI-driven approach offers massive efficiency gains. Manual quantitative analysis typically takes 15-20 minutes per patient, whereas the DNN processed the entire study dataset in roughly 13 minutes—a 90-94% reduction in analysis time. This speed makes real-time quantitative assessment feasible in busy clinical settings.
Conclusion and Limitations
This novel model provides a standardized, reproducible method for risk stratification, allowing for tighter clinical monitoring of vulnerable patients. However, as a single-center study, further multi-center validation is required to ensure generalizability. Future work will focus on achieving 100% automation—as roughly 7.5% of frames currently require manual expert correction—and incorporating additional clinical parameters like pre-PCI flow to further refine predictive accuracy.
Summary updated