Background. The choice of treatment strategy for coronary artery disease is often based on: 1) anatomical information on stenosis locations, and 2) functional information on their haemodynamic relevance, e.g. myocardial deformation or perfusion. Inspecting a single fused image containing both anatomical and functional information, as opposed to viewing separate images side-by-side, facilitates this treatment choice. The aim of this study is to develop a novel cardiac fusion imaging technique to combine 3D+time echocardiography (3DE) (functional information) with coronary computed tomography angiography (CCTA) (anatomical information).
Method. 3DE and CCTA data sets were obtained from 20 patients with suspected coronary artery disease. The coronary artery tree was segmented from the CCTA images. A semi-automatic fusion algorithm was developed to perform the following steps: The left ventricle (LV) 3D surfaces were segmented in the CCTA image and 3DE images and used to align the two data sets. The moving 3DE LV was then visualized along with the CCTA coronary arteries. Myocardial strain was estimated and visualized on the LV surface.
Results. Preliminary fusion results from images of one patient have been obtained. The figure shows the CCTA coronary artery tree aligned with a) 3DE LV endocardium in end-systole, b) 3DE LV endocardium in end-diastole, and c) 3DE LV with colour-coded instantaneous longitudinal strain.
Discussion. Preliminary results show that fusion of CCTA and 3DE images is feasible. However, the algorithm needs to be further developed to increase automation and include other functional parameters, such as myocardial perfusion. Moreover, a validation study to assess algorithm performance and diagnostic value in multiple patients will be performed.