Noninvasive Reconstruction Of Cardiac Electrophysiology Based On Non-local Features

Non-invasive reconstruction of electrophysiological activity in the heart is of great significance for clinical disease prevention and surgical treatment.To obtain 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.Based on the nonlocal self-similarity of TMP distribution,we propose two novel methods to reconstruct the dynamic distribution of TMPThe first one is based on the traditional optimization method.The existing state-of-art method Total Variation Minimization(TV)constraints only take 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,we introduce a novel regularization method called Graph Based Total Variation to make up for the above shortcomings.The graph structure takes the TMP value of a time series on each cardiac node as the criterion to establish the similarity relationship among the heart.Two groups of phantom experiments were set to verify the superiority of the proposed method over the traditional constraints:infarct scar reconstruction and activation wave fronts reconstruction.In addition,experiments with 10 real premature ventricular contractions(PVCs)patient data were used to demonstrate the accuracy of our method in clinical applicationsThe second method combines deep learning technology.Combining the physical interpretation and convergence theory provided by the optimization method with the advantages of deep learning to automatically extract features from big data,we propose a neural network based on Iterative Shrinkage Threshold Algorithm(ISTA-Net),and add a nonlocal feature extraction module.Through experiments,the spatial distribution,time distribution and activation sequence diagram of TMP sequences are analyzed,and the advantages of this method are illustrated.