Computational Analysis of Post-CABG Hemodynamics

The two Aims of this work are summarized in Fig 1. In Aim 1, we performed multiscale CFD simulations of patient-specific post-CABG models (including the native coronary artery stenoses) under computationally hyperemic conditions. Then, we compared the computational FFR in the grafted LAD derived from CFD with the angiographic FFR (CAAS vFFR, Pie Medical Imaging, Maastricht, Netherlands). The angiographic FFR was used as a surrogate for the invasive wire-based FFR [25 (link)]. In Aim 2(i), we 3D reconstructed post-CABG anatomies from CCTA and computationally created focal lumen stenoses with four degrees of severity (mild, moderate, severe, critical) in the LAD (proximal to the LIMA anastomosis; Fig 2). The stenosis of the LAD arteries of Aim 2 was less than 70%, thus exerting minimal impact. Then, in each post-CABG model with the computationally created LAD stenosis, we computationally removed all bypass grafts to reproduce the pre-CABG coronary artery anatomy (Fig 2). Using the computational framework of Aim 1, we compared the local hemodynamics [resting distal coronary pressure to aortic pressure ratio (Pd/Pa) and FFR] in the LAD before and after the LIMA grafting under computational resting and hyperemic conditions. In Aim 2(ii), we used the same post-CABG models as in Aim 2(i) and studied the impact of four different degrees of LAD stenosis on the flow in the LAD and LIMA under computational resting and hyperemic conditions. In total, in each patient of Aim 2, we computationally simulated 16 different conditions, i.e., 4 degrees of stenosis severity (mild, moderate, severe, critical) x 2 hemodynamic conditions (rest and hyperemia) x 2 CABG conditions (pre-CABG and post-CABG).