Non‐invasive reconstruction of dynamic myocardial transmembrane potential with graph‐based total variation constraints

Abstract

Non-invasive reconstruction of electrophysiological activity in the heart is of great significance for clinical disease prevention and surgical treatment. The distribution of transmembrane potential (TMP) in three-dimensional myocardium can help us diagnose heart diseases such as myocardial ischemia and ectopic pacing. However, the problem of solving TMP is ill-posed, and appropriate constraints need to be added. The existing state-of-art method total variation minimisation only takes advantage of the local similarity in space, which has the problem of over-smoothing, and fails to take into account the relationship among frames in the dynamic TMP sequence. In this work, the authors introduce a novel regularisation method called graph-based total variation to make up for the above shortcomings. The graph structure takes the TMP value of a time sequence on each heart node as the criterion to establish the similarity relationship among the heart. Two sets of phantom experiments were set to verify the superiority of the proposed method over the traditional constraints: infarct scar reconstruction and activation wavefront reconstruction. In addition, experiments with ten real premature ventricular contractions patient data were used to demonstrate the accuracy of the authors' method in clinical applications.

Neural network aided extended Kalman filtering for inverse imaging of cardiac transmembrane potential

“…In order to verify the effectiveness of the proposed EKFNet for TMP reconstruction, we conducted experiments in four aspects.1) Performance study: We tested the performance of EKFNet in a cardiac ectopic pacing task and a myocardial infarction detection task (van Dam et al 2009), as the former leads to abnormal conduction of TMP. In contrast, the latter leads to localized non-conduction of TMP.2) Comparative study: In order to show the superiority of EKFNet, we compared it with Tikhonov regularization (Greensite and Huiskamp 1998), TV (Xie et al 2019) and VAENet (Ghimire et al 2018), and KFNet (Huang et al 2022). In addition, qualitative and quantitative analyses were also investigated to demonstrate the advantages of EKFNet.3) Ablation study: to investigate the effect of data-driven Kalman coefficient learning, we compared the performance of models without Kalman coefficient learning (STNet).…”

“…Tikhonov is the most classical regularization method for the TMP reconstruction inverse problem (Hofmann et al 2007), and the basic idea is to add a regularization term to the objective function to penalize the complexity of the model. Jingjia et al (2013Jingjia et al ( , 2014 proposed a new total variation (TV) method to exploit the unique spatial properties of segmental smoothing of transmembrane action potentials to improve the robustness of mesh resolution (Xie et al 2019). introduced a graph operator on top of the TV to establish the similarity relationship between hearts using the TMP value of the time series on each heart node as a criterion.…”

Nonlocal based FISTA network for noninvasive cardiac transmembrane potential imaging

“…Some approaches make use of temporal constraints, such as singular value decomposition (SVD) (Greensite 1996), while the other methods, i.e. graph-based TV (Xie et al 2019) and model-based general Bayesian framework (Wang , Jiang et al 2022a) are proposed in order to take into account the spatial and temporal correlation. There are also novel efforts (Aydn andSerinagaoglu 2009, Erem et al 2014) combining nonlocal features or appropriately using the dynamic information of the entire sequences.…”