Fusion of morphological data obtained by coronary computed tomography angiography with quantitative echocardiographic data on regional myocardial function

Abstract

Background: Three-dimensional (3D) fusion of morphological data obtained by coronary computed tomography angiography (CCTA) with functional data from resting and stress echocardiography could potentially provide additional information compared to examination results analyzed separately and increase the diagnostic and prognostic value of non-invasive imaging in patients with suspected coronary artery disease (CAD). Using vendor-independent software developed in our institution, we aimed to assess the feasibility and reproducibility of 3D fusion of morphological CCTA data with echocardiographic data regarding regional myocardial function.

Methods: Thirty patients with suspected CAD underwent CCTA and resting transthoracic echocardiography. From CCTA we obtained 3D reconstructions of coronary arteries and left ventricle (LV). Offline speckle-tracking analysis of the echocardiographic images provided parametric maps depicting myocardial longitudinal strain in 17 segments of the LV. Using our software, 3 independent investigators fused echocardiographic maps with CCTA reconstruc­tions in all patients. Based on the obtained fused models, each segment of the LV was assigned to one of the major coronary artery branches.

Results: Mean time necessary for data fusion was 65 ± 7 s. Complete agreement between independent investigators in assignment of LV segments to coronary branches was obtained in 94% of the segments. The average coefficient of agreement (kappa) between the investigators was 0.950 and the intra-class correlation coefficient was 0.9329 (95% CI 0.9227–0.9420).

Conclusions: Three-dimensional fusion of morphological CCTA data with quantitative echocardiographic data on regional myocardial function is feasible and allows highly repro­ducible assignment of myocardial segments to coronary artery branches.