Tailored Calibration Approach for Hemodynamic Boundary Conditions in CT-FFR

Area of Science:

• Cardiovascular imaging and intervention
• Computational fluid dynamics in cardiology
• Artificial intelligence in medical diagnostics

Background:

• Computed tomography-derived fractional flow reserve (CT-FFR) traditionally uses standardized boundary conditions.
• These standardized conditions do not account for individual patient physiology.
• A novel calibration method is needed to personalize CT-FFR.

Purpose of the Study:

• To develop and assess a calibration method for individualizing CT-FFR boundary conditions using invasive fractional flow reserve (FFR) measurements.
• To evaluate the impact of this calibration on predicting lesion-specific pressure drops (ΔFFR), post-percutaneous coronary intervention (PCI) FFR, and high-risk plaque (HRP) discrimination.

Main Methods:

• Retrospective analysis of 79 patients with single left anterior descending lesions.
• Lumped parameter models were employed to calibrate boundary conditions by minimizing discrepancies between invasive FFR and CT-FFR.
• High-risk plaque (HRP) was identified using an AI system; discrimination was assessed via ΔCT-FFR and CT-FFR gradient.

Main Results:

• Calibration significantly improved FFR correlation from 0.590 to 0.997 (p<0.001) and ΔFFR correlation from 0.490 to 0.823 (p<0.001).
• Post-PCI validation demonstrated improved correlation from 0.440 to 0.871 (p=0.002).
• HRP discrimination improved for both ΔCT-FFR (AUC 0.62 to 0.70, p=0.018) and CT-FFR gradient (AUC 0.62 to 0.69, p=0.032).

Conclusions:

• Individualized boundary condition calibration using invasive FFR allows for accurate hemodynamic prediction.
• This approach enhances the prediction of post-intervention physiology.
• The method shows promise for improved procedural planning and risk stratification in coronary angiography.