Electrical impedance tomography (EIT) for quantification of pulmonary edema in acute lung injury

Abstract

Background: Assessment of pulmonary edema is a key factor in monitoring and guidance of therapy in critically ill patients. To date, methods available at the bedside for estimating the physiologic correlate of pulmonary edema, extravascular lung water, often are unreliable or require invasive measurements. The aim of the present study was to develop a novel approach to reliably assess extravascular lung water by making use of the functional imaging capabilities of electrical impedance tomography.

Methods: Thirty domestic pigs were anesthetized and randomized to three different groups. Group 1 was a sham group with no lung injury. Group 2 had acute lung injury induced by saline lavage. Group 3 had vascular lung injury induced by intravenous injection of oleic acid. A novel, noninvasive technique using changes in thoracic electrical impedance with lateral body rotation was used to measure a new metric, the lung water ratioEIT, which reflects total extravascular lung water. The lung water ratioEIT was compared with postmortem gravimetric lung water analysis and transcardiopulmonary thermodilution measurements.

Results: A significant correlation was found between extravascular lung water as measured by postmortem gravimetric analysis and electrical impedance tomography (r = 0.80; p < 0.05). Significant changes after lung injury were found in groups 2 and 3 in extravascular lung water derived from transcardiopulmonary thermodilution as well as in measurements derived by lung water ratioEIT.

Conclusions: Extravascular lung water could be determined noninvasively by assessing characteristic changes observed on electrical impedance tomograms during lateral body rotation. The novel lung water ratioEIT holds promise to become a noninvasive bedside measure of pulmonary edema.

Fig. 1

Process for extracting the tidal ventilation of the left lung and the tidal ventilation of the right lung (TV_L and TV_R, respectively) from sets of electrical impedance tomographic images. RMS root mean square

Fig. 2

Calculation of lung water ratio_EIT: Trend slopes (TS) computed from the imbalance coefficient (IM_L–R) of each pig in the three different body positions of the protocol. a Conditions in healthy lungs. b Conditions in injured lungs

Fig. 3

Box plots for lung water ratio_EIT. Data for the three study groups before and after induction of experimental lung injury are presented. *Statistically significant difference compared with baseline before lung injury (p < 0.05)

Fig. 4

Correlation of lung water ratio_EIT after induction of experimental lung injury and extravascular lung water (EVLW) by postmortem gravimetry