Electrical impedance tomography for non-invasive assessment of stroke volume variation in health and experimental lung injury

Abstract

Background. Functional imaging by thoracic electrical impedance tomography (EIT) is a non-invasive approach to continuously assess central stroke volume variation (SVV) for guiding fluid therapy. The early available data were from healthy lungs without injury-related changes in thoracic impedance as a potentially influencing factor. The aim of this study was to evaluate SVV measured by EIT (SVVEIT) against SVV from pulse contour analysis (SVVPC) in an experimental animal model of acute lung injury at different lung volumes. Methods. We conducted a randomized controlled trial in 30 anaesthetized domestic pigs. SVVEIT was calculated automatically analysing heart-lung interactions in a set of pixels representing the aorta. Each initial analysis was performed automatically and unsupervised using predefined frequency domain algorithms that had not previously been used in the study population. After baseline measurements in normal lung conditions, lung injury was induced either by repeated bronchoalveolar lavage (n=15) or by intravenous administration of oleic acid (n=15) and SVVEIT was remeasured. Results. The protocol was completed in 28 animals. A total of 123 pairs of SVV measurements were acquired. Correlation coefficients (r) between SVVEIT and SVVPC were 0.77 in healthy lungs, 0.84 after broncho-alveolar lavage, and 0.48 after lung injury from oleic acid. Conclusions. EIT provides automated calculation of a dynamic preload index of fluid responsiveness (SVVEIT) that is non-invasively derived from a central haemodynamic signal. However, alterations in thoracic impedance induced by lung injury influence this method.