Influence of current injection scheme on electrical impedance tomography for monitoring of the respiratory function of obese subjects

Abstract

Electrical impedance tomography (EIT) has been proposed as a noninvasive radiation-free functional imaging modality for bedside monitoring of lung ventilation. Due to the intrinsic low sensitivity of measured boundary voltages with respect to internal conductivity changes, EIT image reconstructions are more challenging for the regions that are located deeper from the attached surface electrodes. Considering the worldwide prevalence of obesity, it is highly desirable for an EIT system to produce high-quality images from obese as well as lean subjects. In EIT, the internal current density distribution and the sensitivity of a boundary voltage to a change of internal conductivity vary with respect to the current injection scheme. We quantified the range of measured boundary voltages subject to a chosen current injection scheme and investigated the influence of various obesity levels on the number of distinguishable boundary voltages given a fixed amount of system noise. We found that the 0-skip current injection scheme, where neighboring electrode pairs are used for current injections, could be most favorable to improve spatial resolution but produces the smallest number of distinguishable boundary voltages. To obtain an enough number of distinguishable boundary voltages from an obese subject, the gap between current injection electrodes should be increased depending on the circumference of a chosen imaging plane. We suggested a guideline for current injection schemes in EIT to produce images with better quality from obese as well as lean subjects.