Analysis of the electrical patterns and structural remodeling in atrial fibrillation

Catheter ablation of atrial fibrillation (AF) is a promising therapy, whose success is limited by uncertainty in the knowledge of the mechanisms sustaining the arrhythmia. Many theories based on atrial electrical activation or on atrial structural remodeling have been proposed to target AF mechanisms. We hypothesized two prospective approaches could be linked and both computational analysis of atrial electrical patterns and fibrotic tissue location and extent could give further insights on the role of rotors and spatial relationship between them and atrial fibrosis. This paper presents some preliminary results aimed at the integration of information derived from electrical patterns and structural remodeling in AF patients. Electrical patterns were analyzed by applying the standard procedure based on the Hilbert transform (HT) and with sinusoidal wavelet recomposition (SR). In addition, a new technique based on the detection of maximum negative derivative of the unipolar electrograms and a modified version of signal recomposition (NDSR) was tested. A patient-specific anatomical model was derived by segmenting magnetic resonance angiographic (MRA) data applying an edgebased level set approach guided by a phase-based edge detector. A multimodality affine registration was applied to register MRA and delayed-enhanced MR imaging (DE-MRI). Following this registration step, gray intensity levels from DE-MRI were used as a texture of the 3D model to visualize fibrosis location and quantify its extent. In view of a future integration of electrical activation patterns on the patient-specific anatomical model, detected atrial activation timings (AAT) and derived parameters were validated with manual annotation performed by an expert cardiologist and the atrial model was compared with the anatomical map used to guide the ablation procedure.