Integrated Strategy to Mitigate Motion-Induced Artifacts During Seizures in Electrical Impedance Tomography

Accurate identification of the epileptogenic zone (EZ) is essential for epilepsy patients to achieve successful surgical outcomes. Electrical impedance tomography (EIT) has the potential to enhance the precision of EZ localization. However, motion-induced image artifacts present a considerable challenge to accurate EIT imaging. This study aims to mitigate motion-induced artifacts in EIT. We propose an integrated strategy to address the artifacts stemming from two scenarios associated with the motion: electrode disconnection and position uncertainty. We employ <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Z</i>-score and Pearson’s correlation coefficient (PCC) of original boundary voltage to identify the disconnected electrodes, a whale optimization algorithm (WOA) optimized backpropagation neural network (BPNN) to correct erroneous data induced by disconnected electrodes and particle swarm optimized variational mode decomposition (PSO-VMD) to suppress the motion interference induced by position uncertainty electrodes. The results show that both EIT voltage and original boundary voltage return to their normal level after motion-induced artifacts suppression, resulting in a marked enhancement in EIT imaging quality. The strategy’s efficacy is confirmed through both animal experiments and trials with healthy volunteers. The methodology presented in this article is adept at mitigating motion interference in brain EIT, improving the accuracy and reliability of EZ localization. It also helps to expand the clinical application scenarios of EIT, enabling the monitoring of other diseases that might produce motion-related interference.