Predictive computational framework to provide a digital twin for personalized cardiovascular medicine

Area of Science:

• Cardiovascular research
• Medical imaging analysis
• Interventional cardiology

Background:

• Percutaneous coronary intervention (PCI) requires precise planning for optimal outcomes.
• Current methods for quantitatively simulating coronary interventions lack high precision and effectiveness.
• Accurate prediction of post-PCI fractional flow reserve (FFR) and vessel anatomy is crucial.

Purpose of the Study:

• To develop and validate a virtual coronary intervention (VCI) technique.
• To enable non-invasive physiological and anatomical assessment for PCI planning.
• To quantitatively simulate coronary intervention outcomes.

Main Methods:

• Enrolled patients with substantial coronary lesions (pre-PCI CT-FFR < 0.80).
• Utilized a VCI framework to predict vessel reshaping and post-PCI CT-FFR.
• Validated predicted post-VCI CT-FFR, luminal cross-sectional area (CSA), and centreline curvature against post-PCI CT angiography data.

Main Results:

• VCI simulation time averaged 24.92 seconds.
• Predicted post-VCI CT-FFR (0.90 ± 0.08) closely matched calculated post-PCI CT-FFR (0.92 ± 0.09).
• Morphological predictions (CSA, centreline curvature) showed good agreement with post-PCI measurements.

Conclusions:

• The VCI technique provides non-invasive pre-procedural assessment of coronary intervention.
• The VCI model demonstrates potential to optimize PCI planning.
• This approach may improve the safety and efficiency of PCI procedures.