PLoS One. 2017 Aug 1;12(8):e0182215.

Effects of individualized electrical impedance tomography and image reconstruction settings upon the assessment of regional ventilation distribution: Comparison to 4-dimensional computed tomography in a porcine model

Florian Thürk1, Daniel Mudrak1, Stefan Kampusch1, Alice Wielandner3, Helmut Prosch3, Johannes Hofmanninger3, Christina Braun4, Frédéric P.R. Toemboel2, Stefan Böhme2 and Eugenijus Kaniusas1

1 Institute of Electrodynamics, Microwave and Circuit Engineering, Vienna University of Technology, Gußhausstraße 27-29, 1040 Vienna, Austria

2 Medical University of Vienna, Department of Anesthesia, Pain Management and General Intensive Care Medicine, Vienna, Austria

3 Medical University of Vienna, Department of Biomedical Imaging and Image Guided Therapy, Vienna, Austria

4 University of Veterinary Medicine Vienna, Department of Anesthesiology & Intensive Care, Vienna, Austria



Electrical impedance tomography (EIT) is a promising imaging technique for bedside monitoring of lung function. It is easily applicable, cheap and requires no ionizing radiation, but clinical interpretation of EIT-images is still not standardized. One of the reasons for this is the “ill-posed” nature of EIT, allowing a range of possible images to be produced – rather than a single explicit solution. Thus, to further advance the EIT technology for clinical application, thorough examinations of EIT-image reconstruction settings is essential.

In the present work, regional ventilation distribution profiles derived from different EIT finite-element reconstruction models and settings (for GREIT and Gauss Newton) were compared to regional aeration profiles assessed by the gold-standard of 4-dimensional computed tomography (4DCT). Specifically, non-individualized reconstruction models (based on circular and averaged thoracic contours) and individualized reconstruction models (based on true thoracic contours) were compared. In addition, the effect of the sizes of region of interests (ROI) as well as pulmonary regions were investigated.

Our results suggest that GREIT with noise figure of 0.15 and non-uniform background works best for the assessment of regional ventilation distribution by EIT, as verified versus 4DCT. Furthermore, the root mean square error (RMSE) of anteroposterior ventilation profiles decreased while correlation increased after embedding anatomical information into the reconstruction models. Reducing the resolution by increasing ROIs size showed better correlation.

In conclusion, the present work reveals that anatomically enhanced EIT-image reconstruction is superior to non-individualized reconstruction models, but further investigations in humans, so as to standardize reconstruction settings, is warranted.





Electrical Impedance Tomography (EIT) is a novel imaging modality for continuous lung monitoring at the bed side. Small, imperceptible currents are injected via surface electrodes into the body and the resulting voltages are measured. From these voltage measurements, the impedance changes within the body – mainly originating from air entering and leaving the lungs – can be visualized. Since EIT does not rely on harmful ionizing radiation, like computed tomography (CT), and the required equipment is compact and relatively cheap, it combines the benefits of clinical monitoring (e.g. readily available, continuous measurement, etc.) and standard imaging modalities (i.e. regional information). It has already been shown in validation studies against CT, positron-emission tomography, xenon, and other imaging modalities [1]–[4], that regional ventilation distribution can be assessed by EIT. Such comparisons are often performed by dividing EIT and CT images into equally sized regions of interests (ROI), e.g. quadrants or horizontal ROIs, or by calculating further ventilation parameters, e.g. center of ventilation (CoV).

Even though these experimental evaluations were rather successful in the past, application of EIT in the clinical routine is still limited due to lacking standards for image interpretation and decision pathways [5]. While clinical trials are ongoing to establish the necessary knowledge foundation and rationale, also the underlying image reconstruction techniques have to be further improved and standardized. Since the pathways of current within the body are highly dispersed in all dimensions, the reconstruction of images still represents a major challenge and different reconstruction algorithms yield different images; i.e. the mathematical formulation is “ill-posed” [6]. In order to reduce the range of possible EIT images, forward models, e.g. finite element models (FEM), are often utilized as sensitivity model to simulate boundary (electrode) voltages based on internal conductivity changes. Here, the shape and anatomical accuracy of these FEMs determine the quality of image reconstruction, but the true body shape is often not available. Efforts exist to include average contours from large patient cohorts, but strong inter-patient variability (especially in critical patients) limit their validity. Clinical devices often also use simple geometric thorax shapes, e.g. circles or ellipses, which are known to produce reconstruction errors.

In order to address the issues of EIT-image reconstruction, the aim of the present study was first to identify optimal reconstruction settings for GREIT [7] and Gauss Newton [8] reconstruction algorithms. Secondly, applying these settings, we investigated the impact of different reconstruction models in their ability to assess regional ventilation distribution. Specifically, reconstruction models were based on (1) standard circular shape, (2) averaged thoracic body contours, and (3) individualized thoracic body contours. Moreover, EIT-images derived from (2) and (3) were further enhanced by integration of anatomical lung borders to extract pulmonary EIT-pixels only.


In addition to previous analysis, here, also an evaluation of lung mask sizes and of the optimal ROI size is presented. In addition, physiological parameters right-left ratio and anterior-posterior center of ventilation were investigated for different model geometries (see Fig. 1).



Fig.1. Comparison of CT (left) and EIT (right) tidal images. Normalized tidal changes of air content and impedance change along with the corresponding anterior-posterior ventilation distribution, vd, are shown. Different sizes of region of interests (q4-q32), individual lung contours as well as physiological parameters center of ventilation and right-left ratio are indicated.


In order to analyze the effect of lung masks size, the true lung contours – as acquired from CT images – are gradually decreased and increased in size using morphological image operations. In fact, the correlation of anteroposterior ventilation distribution was highest with the true lung mask. Reducing the ROI size immediately strongly decreased the correlation but the ROI size could be increased by more than 200% before a reduction could be observed (compare Fig. 2a). The lowest correlation was achieved with mask size equal to image size, which corresponds to not using any mask. This is in line with our previous findings. To find the optimal ROI size, CT and EIT images were divided into 4, 8, 16, 32 and 64 (q4 – q64) equally sized quadrants before comparison, as indicated in Fig.1. For comparison, correlation between CT and EIT of anterior-posterior and right-left ventilation profiles were considered. As anticipated, correlations were highest for largest quadrant size (q4) and decrease with increasing resolution. Overall, matching between CT and EIT was lower for right-left compared to anterior-posterior ventilation distribution (compare Fig.2b).



Fig.2. a) Correlation between anterior-posterior ventilation distributions from CT and EIT for different lung contour / mask sizes for pulmonary pixel selection. b) Correlation for anterior-posterior and right-left ventilation distributions after dividing the image into regions of interest of different sizes (q4 … 64 quadrants).


Finally, physiological parameters R:L and CoV were calculated for each pig and each reconstruction model. Values of R:L and CoV varied strongly between reconstruction models and did not match ground truth data from CT. For both parameters, individualized EIT seemed to produce better results than averaged or circular reconstruction models (see Fig. 3).



Fig.3. a) Anterior-posterior center of ventilation and (b) right-left-ratio of images using different reconstruction models.

Circular      … no anatomical information,
Average    … averaged lung and thorax contours
Individual   … individual lung and thorax contours
Average+    … Average with pulmonary pixels only
Individual+ … Individual with pulmonary pixels only


In this work, evaluations were performed in order to quantify the comparability between CT and EIT images with different lung mask sizes and ROI sizes. In addition, the influence of reconstruction models on physiological parameters, CoV and R:L, was analyzed for reconstruction models with different thorax geometries.

Our result highlight the importance of standardized methods for future clinical application of EIT. Using different lung regions or varying the size of ROIs can strongly change the results of an analysis [9]. In the case of unknown true geometries, our results suggest that, generally, choosing larger lung masks produces more robust results. Further, physiological parameters CoV and R:L depend on the geometry of the reconstruction model, and prior information of a bodies anatomy can decrease the error as validated with CT. Future works should therefore focus on establishing a standardized procedure to transfer anatomical contours into clinical EIT image reconstruction.



[1]      H. Wrigge, J. Zinserling, T. Muders, D. Varelmann, U. Gunther, C. von der Groeben, A. Magnusson, G. Hedenstierna, and C. Putensen, “Electrical impedance tomography compared with thoracic computed tomography during a slow inflation maneuver in experimental models of lung injury.,” Crit. Care Med., vol. 36, no. 3, pp. 903–909, 2008.

[2]      J. C. Richard, C. Pouzot, A. Gros, C. Tourevieille, D. Lebars, F. Lavenne, I. Frerichs, C. Guérin, R. M. U, and J. Capelle, “Electrical impedance tomography compared to positron emission tomography for the measurement of regional lung ventilation: an experimental study.,” Crit. Care, vol. 13, no. 3, p. R82, 2009.

[3]      J. Hinz, P. Neumann, T. Dudykevych, L. G. Andersson, H. Wrigge, H. Burchardi, and G. Hedenstierna, “Regional ventilation by electrical impedance tomography: A comparison with ventilation scintigraphy in pigs,” Chest, vol. 124, no. 1, pp. 314–322, 2003.

[4]      G. Elke, M. K. Fuld, A. F. Halaweish, B. Grychtol, N. Weiler, E. A. Hoffman, and I. Frerichs, “Quantification of ventilation distribution in regional lung injury by electrical impedance tomography and xenon computed tomography,” Physiol Meas, vol. 34, no. 10, pp. 1303–1318, 2013.

[5]      I. Frerichs, M. B. P. Amato, A. H. van Kaam, D. G. Tingay, Z. Zhao, B. Grychtol, M. Bodenstein, H. Gagnon, S. H. Böhm, E. Teschner, O. Stenqvist, T. Mauri, V. Torsani, L. Camporota, A. Schibler, G. K. Wolf, D. Gommers, S. Leonhardt, and A. Adler, “Chest electrical impedance tomography examination, data analysis, terminology, clinical use and recommendations: consensus statement of the TRanslational EIT developmeNt stuDy group,” Thorax, p. thoraxjnl-2016-208357, 2016.

[6]      D. Sbarbaro, M. Vauhkonen, and T. a Johansen, “State estimation and inverse problems in electrical impedance tomography: observability, convergence and regularization,” Inverse Probl., vol. 31, no. 4, p. 45004, 2015.

[7]      A. Adler, J. H. Arnold, R. Bayford, A. Borsic, B. Brown, P. Dixon, T. J. C. Faes, I. Frerichs, H. Gagnon, Y. Gärber, B. Grychtol, G. Hahn, W. R. B. Lionheart, A. Malik, R. P. Patterson, J. Stocks, A. Tizzard, N. Weiler, and G. K. Wolf, “GREIT: a unified approach to 2D linear EIT reconstruction of lung images.,” Physiol. Meas., vol. 30, no. 6, pp. S35–S55, 2009.

[8]      T. J. Yorkey, J. G. Webster, and W. J. Tompkins, “Comparing Reconstruction Algorithms for Electrical Impedance Tomography,” IEEE Trans. Biomed. Eng., vol. BME-34, no. 11, pp. 843–852, 1987.

[9]      S. Pulletz, H. R. van Genderingen, G. Schmitz, G. Zick, D. Schädler, J. Scholz, N. Weiler, and I. Frerichs, “Comparison of different methods to define regions of interest for evaluation of regional lung ventilation by EIT.,” Physiol. Meas., vol. 27, no. 5, pp. S115–S127, 2006.