Menu Close

Spatial and Temporal Inverse Problem Techniques for Arrhythmia Mechanism Analysis

Authors

Raúl Paul Caulier Cisterna

Advisor(s)

José Luis Rojo Álvarez y Margarita Sanromán Junquera

Research area

ICT for Healthcare

Affiliated Research Center

Universidad Rey Juan Carlos

In the field of bioengineering, specifically in cardiac electrophysiology research, there is a lot of scientific work on techniques for the early detection, identification and treatment of cardiac arrhythmias, which are one of the leading causes of death in the world. These cardiac arrhythmias are under constant investigation due to the lack of knowledge in their mechanisms and the difficulty in their location and identification. On the other hand, from a cardiological point of view, cardiac arrhythmias are defined as a disorder in the normal heart rhythm, generated naturally in response to physiological needs, or by disturbances in the electrical activity that controls the contraction of the cardiac muscle. These arrhythmias are caused, in most cases, by regions of the heart that are diseased, with slow conduction, or with scars. One of the main information sources for the study of arrhythmias are the intracavitary electrograms (EGMs). An EGM is an endocardial recording of the bioelectrical signal generated by the electrical impulse that contracts the cardiac muscles. The procedure used to register the EGMs, identify the arrhythmia mechanism, and treat it with ablation therapy is the electrophysiological study (EPS). Cardiac ablation consists of the suppression of the arrhythmia by cauterizing the diseased cardiac tissue with radiofrequency or intense cold using intracavitary catheters. As supporting systems in the EPS, cardiac navigation systems (CNS) are used to generate maps with anatomical and electrical information of the cardiac cavity. These maps help to identify the arrhythmia and the sequence of activation by registering the EGMs in different spatial locations of the cardiac chamber. The number of points and their spatial distribution are determined during the procedure heuristically and require a time of exploration and sequential recording of the EGMs that lengthens the duration of the EPS. At present, non-invasive systems for the generation of maps with anatomical and electrical information of the heart have been developed. These systems, called electrocardiographic image (ECGI) systems, estimate the epicardial unipolar EGMs in an cardiac epicardial mesh by using the variation of the electric field generated by the bioelectric currents and registered with electrodes located on the patient torso and back. This allows us to obtain meshes of the epicardial unipolar EGMs with a high number of nodes in real time and in a non-invasive way. The ECGI system uses the inverse problem in electrocardiography to calculate and estimate these virtual EGMs. This inverse problem is a numerically ill-conditioned problem, due to the inversion of the matrix characteristics. Although regularization methods, such as Tikhonov or the Decomposition of Truncated Singular Values, have been proposed to solve this problem, they still present resolution problems in their provided solution. On the other hand, ECGI systems are limited in their use with the EPS criteria, since bipolar EGM are used in EPS while unipolar EGM are provided by ECGI. The width of the EGM complex, the amplitude of the depolarization wave, and the fragmentation of the bipolar EGM are features used in the EPS. These bipolar EGMs in EPS are obtained by the subtraction of the unipolar EGM captured with a pair of catheter electrodes. In the ECGI systems, the unipolar EGM are used to create activation maps of the heart, but this information is insufficient to perform an analysis of the EGMs and to determine the type of arrhythmia or its location.