Multi-parametric model of the heart from CT images to guide ventricular tachycardia ablation

The purpose of this work was to construct a 3D multi-parametric model of the heart by automatically segmenting cardiac cavities, left myocardium, scar and epicardial fat from multidetector computed tomographic (MDCT) volumes, using a level set algorithm based on a new multi-scale stopping function. This method was applied to 4 patients with recurrent ventricular tachycardia (VT) undergoing contrast enhanced (CE)-MDCT imaging, composed by an angiographic (ANGIO) and a late enhanced (LE) scan, before electro-anatomic mapping (EAM) and radiofrequency ablation (RFa). The segmented structures were integrated into the clinical surgery software system (CARTO). The adequacy of our model was verified by an expert radiologist and an arrhythmologist using a qualitative score.