Cardiac resynchronization therapy (CRT) is an effective treatment for patients with congestive heart failure and ventricular dyssynchrony. Despite the overall efficacy of CRT, approximately 30% of patients receiving CRT do not improve. One of the main technical problems related to the CRT procedure is inadequate visualisation in X-ray fluoroscopy of the venous anatomy in relation to accurate cardiac chamber visualisation. This paper proposes a novel approach for 3D reconstruction of coronary veins from a single contrast enhanced intraoperative fluoroscopy image. For this application, the method uses back-projection geometry and a Euclidean distance/angle-based cost function. The algorithm is validated on a phantom and five patient datasets, comprising six view-angle orientations for the phantom dataset and two view-angle orientations for each of the patient datasets. Median(interquartile range) 3D-reconstruction accuracies of 1.41(0.55-3.00) mm and 3.28(2.10-4.89) mm were established for the phantom and patient data, respectively. The technique can facilitate careful advancement of the cannulating guide over a guidewire or a diagnostic catheter positioned in the coronary sinus, and consequently, improve the chances of response to CRT.
CITATION STYLE
Panayiotou, M., Toth, D., Adem, T., Mountney, P., Brost, A., Behar, J. M., … Rhode, K. S. (2017). 3D reconstruction of coronary veins from a single X-ray fluoroscopic image and pre-operative MR. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10124 LNCS, pp. 66–75). Springer Verlag. https://doi.org/10.1007/978-3-319-52718-5_8
Mendeley helps you to discover research relevant for your work.