Separation of Cardiac- and Ventilation-related Signals within Electrical Impedance Tomography Data based on Multi-dimensional Ensemble Empirical Mode Decomposition

Abstract

Electrical impedance tomography (EIT) is an promising imaging technology for continuous bedside monitoring of ventilation and perfusion. However, due to the spatial and frequency overlapping of ventilation and cardiac components in the heart-lung interaction system, it’s difficult to separate the components in spontaneous breathing subjects. We introduce an intuitive method based on multi-dimensional ensemble empirical mode decomposition to explore the intrinsic oscillation modes of the ventilation and cardiac components from EIT data. This study combines the spatial information with temporal information, and establishes the combination strategy for the two physiological components based on multi-scale analysis. Our study illustrates preliminary in-vivo results based on the data collected from two healthy male subjects, and qualitatively validates the efficiency of resolving the overlapping of ventilation and perfusion component. The method proposed in our study is believed to open up new possibilities for the assessment of lung ventilation and perfusion. In future work, quantitative validation for separation results of ventilation component and perfusion component will be conducted.