RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación
Artículo científico (JournalArticle)
Electrocardiographic imaging including intracardiac information to achieve accurate global mapping during atrial fibrillation
[EN] Atrial fibrillation (AF) is characterized by complex and irregular propagation patterns. Noninvasive electrocardiographic imaging (ECGI) has been tested during AF conditions with promising results. However, current regularization methods face important challenges in this type of unstable electrical activity scenarios. Combination of intracardiac and non-invasive simultaneous recordings could improve ECGI performance and allow real-time global mapping of complex AF patterns. In this work, we propose an ECGI method that incorporates intracardiac measurements as a constraint in a reformulation of the classical Tikhonov method. We used realistic mathematical models of atria and torso that simulates a wide number of epicardial electrical activity patterns. Body surface potentials were obtained from simulated electrograms (EGMs) by using Boundary Element Method and corrupted with Gaussian noise. Epicardial potentials were estimated using inverse problem with Tikhonov regularization, including intracavitary information as a second constraint. Results showed that first-order Constrained Tikhonov formulation provided more reliable reconstructions than the classical Tikhonov approach in AF conditions using at least 32 uniformly distributed endocardial EGMs (CC between 0.87 and 0.28, depending on the AF complexity). Constrained Tikhonov provided more accurate spatial mass functions (SMF) of PS locations (CC smF between 0.24 and 0.86). This methodology was tested on real patient data, obtaining a mean DF RMSE of 0.85 Hz, outperforming the classical Tikhonov approach. Limitations of this study include the fact that the model considered endocardium and epicardium as a single layer. Further research will include endocardium-epicardium bilayer model approximations and validation using more real patient data., This work has been partially supported by projects TEC2016-75361R and PID2019-105032GB-I00 from the Spanish Ministry of Economy and Spanish Ministry of Science and Innovation, respectively, projects IJCI-2014-22178, PI16-01123, DTS16/00160, PI17/0159, PI17/01106 from the Spanish Ministry of Economy through the Carlos III Health Institute with FEDER founds, grant GVA APOSTD/2017/068 and projects AICO/2018/267, GV/2018/103 from the Education, Research, Culture and Sports department of Generalitat Valenciana, Spain.
Proyecto:
Agencia Estatal de Investigación / Plan Estatal de Investigación Científica y Técnica y de Innovación 2017-2020 PID2019-105032GB-I00
,
Ministerio de Economía y Competitividad TEC2016-75361-R
,
Ministerio de Economía y Competitividad DTS16/00160
,
Instituto de Salud Carlos III PI17/01059
,
Ministerio de Economía y Competitividad IJCI-2014-22178
,
Instituto de Salud Carlos III PI17/01106
,
Ministerio de Economía y Competitividad PI16/01123
,
Generalitat Valenciana APOSTD/2017/068
,
Generalitat Valenciana GV/2018/103
,
Generalitat Valenciana AICO/2018/267