Comparison of Left Ventricular Late Activation Zones Determined Using Noninvasive Electrocardiographic Imaging with Sequential Computed and Magnetic Resonance Tomography

Abstract

Recently, noninvasive electrocardiographic imaging (ECGI) has been introduced into the clinical practice. Although the ECGI diagnostic accuracy in combination with computed tomography (CT) was previously reported, the question of validating this technique with use of magnetic resonance imaging (MRI) was not solved. In this article, the method of comparison between left ventricular late activation zones determined by ECGI and sequential CT and MRI is presented. To demonstrate performance of the proposed methodology, we studied eleven patients before implantation devices for cardiac resynchronization therapy. The reconstructed three-dimensional CT and MRI models of the ventricles were merged using custom software written in Python. The median (LQ; UQ) mismatch in the obtained late activation zones was 12 (7; 13) mm. Minimal distance was 1 mm, while the maximum error was 20 mm. To conclude, this study is the first attempt to validate the accuracy of ECGI in combination with MRI against the CT-based gold standard in determining the late activation zones.