Artificial Intelligence-Enabled Quantitative Coronary Plaque and Hemodynamic Analysis for Predicting Acute Coronary...

Area of Science:

• Cardiology
• Medical Imaging
• Artificial Intelligence

Background:

• Accurate risk prediction for acute coronary syndrome (ACS) requires improved characterization of coronary lesions.
• Current methods for assessing coronary lesions lack sufficient detail for precise risk stratification.

Purpose of the Study:

• To evaluate the added value of artificial intelligence-enabled quantitative coronary plaque and hemodynamic analysis (AI-QCPHA) in predicting ACS culprit lesions.
• To compare the predictive performance of a model incorporating AI-QCPHA features against a conventional risk model.

Main Methods:

• Retrospective analysis of ACS patients who underwent coronary computed tomography angiography (CTA) prior to their event.
• Culprit and non-culprit lesions were identified and adjudicated using invasive coronary angiography.
• A reference model (CT-derived stenosis severity and plaque characteristics) was compared with a new model including AI-QCPHA features (e.g., fractional flow reserve, plaque burden, blood flow).
• Model performance was assessed using area under the curve (AUC) in a validation cohort.

Main Results:

• The study included 351 patients with 2,088 non-culprit and 363 culprit lesions.
• AI-QCPHA identified key features such as lesion fractional flow reserve, plaque burden, and myocardial blood flow.
• The AI-QCPHA model demonstrated significantly higher predictability for ACS culprit lesions in the validation cohort (AUC 0.84) compared to the reference model (AUC 0.78).
• The enhanced predictability was consistent across various time intervals between CTA and the ACS event.

Conclusions:

• AI-enabled quantitative analysis of coronary plaque and hemodynamics significantly enhances the predictability of ACS culprit lesions.
• This AI-QCPHA approach provides a more accurate risk stratification tool compared to conventional coronary CTA analysis.
• The findings support the integration of AI-QCPHA into clinical practice for improved ACS risk assessment.