Pranav Bhagirath

180 Chapter 9 A-priori model In literature [20–22], inverse procedures involving 64 or fewer electrodes always apply an a priori activation model. Although detailed images may be obtained using even <20 electrodes, the number of degrees of freedom in this situation fundamentally limits the number of pathology related activation patterns that can be represented. This is further elaborated upon in appendix and in figure 7 . Figure 7. Linear mapping M denotes the ideal reconstruction, reflecting the unknown but exact linear quasi-stationary field equations. M maps n independent electrode signals onto at most n independent epicardial base vectors. Note that, by definition of M, no reconstruction method can do better. In practice, ill conditioning will render certain subsets of epicardial patterns indistinguishable, reducing the dimension of the solution space. Obtaining detailed images of cardiac surface potentials using a limited number of electrodes, without excluding a wide variety of pathological activation patterns by constraining the solutions using an a priori activation model derived from healthy myocardial tissue, requires focusing all degrees of freedom on a limited region of interest on the cardiac surface. MRI Although computed tomography (CT) is frequently preferred due to the speed of the acquisition process, MRI allows reliable function analysis, assessment of wall motion abnormalities and highly detailed characterisation of tissue [23–26]. Unlike CT, MRI does not use radiation. Hence, MRI is the preferred imaging modality to be repeatedly used in patients. In recent years, the safe performance of MRI in patients with non-MRI- conditional pacing devices has been demonstrated [27–29]. In addition, MRI-conditional devices have been introduced, decreasing the risk of potential hardware or software interactions [30].