Source Consistency Electrical Impedance Tomography

Abstract

In electrical impedance tomography (EIT), multiple electrodes are attached around an imaging domain such as the human thorax to inject currents and measure induced boundary voltages. Using the measured boundary voltage data, cross-sectional images of an internal conductivity distribution are reconstructed. Taking advantage of its fast temporal resolution, time-difference EIT can be used for image-based monitoring of physiological functions such as lung ventilation and cardiac blood flow. Among numerous data collection protocols, we assume current injections and voltage measurements between adjacent pairs of electrodes. The measured voltage difference between the $j$th electrode pair subject to the current injection between the $k$th electrode pair, for example, changes with time and its time-series is called a voltage channel in this paper. Investigating shapes of voltage channels, a new technique called source consistency EIT (scEIT) is proposed to extract voltage channel data originating from a physiological function or source of interest. The proposed scEIT technique suggests each voltage channel can be expressed up to a scale factor and offset value from a single shape-reference voltage channel when there exists only one time-varying source. When multiple physiological sources exist to concurrently produce correspondingly different conductivity changes, measured voltage channels are influenced by all of the sources. Using the scEIT method, each voltage channel can be expressed as a weighted sum of multiple shape-reference voltage channels of the sources. The proposed scEIT technique is verified through numerical simulations and animal experiments. Future experimental studies of applying the scEIT technique to in vivo human experiments are proposed.