Lumped-parameter model as a non-invasive tool to assess coronary blood flow in AAOCA patients

Abstract

Anomalous aortic origin of the coronary artery (AAOCA) is a rare disease associated with sudden cardiac death, usually related to physical effort in young people. Clinical routine tests fail to assess the ischemic risk, calling for novel diagnostic approaches. To this aim, some recent studies propose to assess the coronary blood flow (CBF) in AAOCA by computational simulations but they are limited by the use of data from literature retrieved from normal subjects. To overcome this limitation and obtain a reliable assessment of CBF, we developed a fully patient-specific lumped parameter model based on clinical imaging and in-vivo data retrieved during invasive coronary functional assessment of subjects with AAOCA. In such a way, we can estimate the CBF replicating the two hemodynamic conditions in-vivo analyzed. The model can mimic the effective coronary behavior with high accuracy and could be a valuable tool to quantify CBF in AAOCA. It represents the first step required to move toward a future clinical application with the aim of improving patient care. The study was registered at Clinicaltrial.gov with (ID: NCT05159791, date 2021-12-16).

Figure 1

Schematic representation of the lumped-parameter model and input data used to calibrate it. On the top, IVUS registration was illustrated in the yellow rectangle: a slice was selected to extract the geometrical measurements related to the more proximal tract (OSTIUM) of a right CA. In the green rectangle, the points selected along the centerline of the 3D model of the aortic root were shown; they were used to measure aortic geometrical parameters. The 3D model was reconstructed through the segmentation of CT. In the middle, signals recorded during the coronary functional assessment were shown in the blue rectangle. For both rest and hyperemic conditions CBF (Q) and CA resistance (R) were assessed. Aortic pressure wave signal was used to extract patient values to rescale both aortic and left ventricular pressure waves taken from literature. On the bottom, the purple rectangle enclosed a schematic representation of the lumped-parameter model. Each arterial segment was represented by the circuit highlighted in gray, aortic BCs were defined by the Windkessel circuit colored in orange, and the BC of the CAs were defined in the blue dashed rectangle.

Figure 2

Results of the model and comparison with in-vivo data. (a) In-silico computed CBF and in-vivo measured CBF are shown in the boxplots; only significant differences are highlighted (*). (b) Scatter plots and correlation analysis are shown. The results related to the left CA are presented on the left, and those corresponding to the right CA are reported on the right. Left CA results are represented in yellow and orange for rest and hyperemic conditions, respectively. Right CA results are presented in green and blue for rest and hyperemic conditions, respectively. (c) Bland-Altman plot analysis was shown for both left and right CA. Difference between in-silico and in-vivo CBF were reported on y-axis, the average was reported on the x-axis. Horizontal dashed lines represented the the 95$\%$ limits of agreement of rest (docs) and hyperemic (doc and line) condition. The blue and red lines represented the average discrepancy between in-silico and in-vivo measurements for rest and hyperemia, respectively.