Recent Advances in and Limitations of Cardiac Output Monitoring by Means of Electrical Impedance Tomography

Pikkemaat, Robert Dipl.-Ing*; Lundin, Stefan MD, PhD; Stenqvist, Ola MD, PhD; Hilgers, Ralf-Dieter Dr. rer. Nat.; Leonhardt, Steffen Dr. med. Dr.-Ing.*

doi: 10.1213/ANE.0000000000000241
Technology, Computing, and Simulation: Research Report

BACKGROUND: Currently, the monitoring of cardiac output (CO) and stroke volume (SV) is mainly performed using invasive techniques. Therefore, performing CO monitoring noninvasively by means of electrical impedance tomography (EIT) would be advantageous for intensive care. Our hypothesis was that, by means of EIT, it is possible to assess heart rate (HR) and to quantify changes in SV due to changes in ventilator settings.

METHODS: CO (HR and SV) of 14 pigs (32–40 kg body weight) was changed by incremental increases in positive end-expiratory pressure levels (0, 5, 10, 15, and 20 cm·H2O; ramp maneuver). This ramp maneuver was applied 4 times in each animal, yielding 43 evaluable single experiments. At each positive end-expiratory pressure level, SV was assessed by transpulmonary thermodilution using a PiCCO device. EIT data were acquired using a Dräger EIT Evaluation Kit 2.

RESULTS: The EIT-based SV-related signal, ZSV (in [AU]), showed only a weak correlation (after excluding 2 measurements) with SVTTD of r = 0.58 (95% confidence interval, 0.43–0.71). If ZSV is calibrated by the reference 1 time for each experiment (defined as SVEIT), the correlation is approximately 0.85 (95% confidence interval, 0.78–0.90). A possible reason for the moderate correlation is the unexpected scaling pattern, leading to amplification of the cardiac impedance signal, found in some animals. The scaling is probably due to the imperfect reconstruction (i.e., a change of sensitivity) of the EIT images or to a change in the position of the heart.

CONCLUSIONS: The hypothesis that EIT can be used to monitor CO and SV was confirmed, but further studies are required before this technique can be applied in clinical practice. HR was determined robustly and accurately. For SV monitoring, promising results were obtained in 80% of the experiments. However, unexpected scaling of the cardiac EIT signal causing inaccurate estimation of SV remains an issue. Before robust assessment of SV by EIT is suitable for clinical practice, the cause of and compensation for undesired scaling effects need to be investigated.

Published ahead of print May 7, 2014

From the *Institute for Medical Information Technology, RWTH Aachen University, Aachen, Germany; Department of Anesthesia and Intensive Care, Sahlgrenska University Hospital Gothenburg, Gothenburg, Sweden; and Department of Medical Statistics, RWTH Aachen University, Aachen, Germany.

Accepted for publication January 23, 2014.

Published ahead of print May 7, 2014

Funding: Dräger Medical GmbH, Lübeck, Germany financed parts of the study.

Conflicts of Interest: See Disclosures at the end of the article.

Reprints will not be available from the authors.

Address correspondence to Robert Pikkemaat, Dipl.-Ing, Institute for Medical Information Technology, RWTH Aachen University, Pauwelsstr. 20, 52074 Aachen, Germany. Address e-mail to

© 2014 International Anesthesia Research Society