Influence of heart motion on cardiac output estimation by means of electrical impedance tomography: a case study. (26th May 2015)

• Record Type: Journal Article
• Authors: Proença, Martin; Braun, Fabian; Rapin, Michael; Solà, Josep; Adler, Andy; Grychtol, Bartłomiej; Bohm, Stephan H; Lemay, Mathieu; Thiran, Jean-Philippe
• Is Part Of: Physiological measurement. Volume 36:Number 6(2015:Jun.)
• Page Start: 1075
• Page End: 1091
• Publication Date: 2015-05-26
• Subjects: electrical impedance tomography -- EIT -- stroke volume -- cardiac output -- non-invasive -- continuous monitoring; Physiology -- Measurement -- Periodicals; Patient monitoring -- Periodicals; 612
• DOI: 10.1088/0967-3334/36/6/1075
• Languages: English
• ISSNs: 0967-3334

Abstract

Abstract: Electrical impedance tomography (EIT) is a non-invasive imaging technique that can measure cardiac-related intra-thoracic impedance changes. EIT-based cardiac output estimation relies on the assumption that the amplitude of the impedance change in the ventricular region is representative of stroke volume (SV). However, other factors such as heart motion can significantly affect this ventricular impedance change. In the present case study, a magnetic resonance imaging-based dynamic bio-impedance model fitting the morphology of a single male subject was built. Simulations were performed to evaluate the contribution of heart motion and its influence on EIT-based SV estimation. Myocardial deformation was found to be the main contributor to the ventricular impedance change (56%). However, motion-induced impedance changes showed a strong correlation ( r = 0.978) with left ventricular volume. We explained this by the quasi-incompressibility of blood and myocardium. As a result, EIT achieved excellent accuracy in estimating a wide range of simulated SV values (error distribution of 0.57 ± 2.19 ml (1.02 ± 2.62%) and correlation of r = 0.996 after a two-point calibration was applied to convert impedance values to millilitres). As the model was based on one single subject, the strong correlation found between motion-induced changes and ventricular volume remains to be verified in larger datasets.