Lung regions identified with CT improve the value of global inhomogeneity index measured with electrical impedance tomography

Lin Yang; Meng Dai; Knut Möller; Inéz Frerichs; Andy Adler; Feng Fu; Zhanqi Zhao

doi:10.21037/qims-20-682

Lung regions identified with CT improve the value of global inhomogeneity index measured with electrical impedance tomography

Quant Imaging Med Surg. 2021 Apr;11(4):1209-1219. doi: 10.21037/qims-20-682.

Authors

Lin Yang¹, Meng Dai², Knut Möller³, Inéz Frerichs⁴, Andy Adler⁵, Feng Fu², Zhanqi Zhao^{2

3}

Affiliations

¹ Department of Aerospace Medicine, Fourth Military Medical University, Xi'an, China.
² Department of Biomedical Engineering, Fourth Military Medical University, Xi'an, China.
³ Institute of Technical Medicine, Furtwangen University, Villingen-Schwenningen, Germany.
⁴ Department of Anaesthesiology and Intensive Care Medicine, University Medical Centre of Schleswig-Holstein Campus Kiel, Germany.
⁵ Department of Systems and Computer Engineering, Carleton University, Ottawa, Canada.

Abstract

Background: The global inhomogeneity (GI) index is a functional electrical impedance tomography (EIT) parameter which is used clinically to assess ventilation distribution. However, GI may underestimate the actual heterogeneity when the size of lung regions is underestimated. We propose a novel method to use anatomical information to correct the GI index calculation.

Methods: EIT measurements were performed at the level of the fifth intercostal space in six patients with acute respiratory distress syndrome. The thorax and lungs were segmented automatically from serial individual CT scans. The anatomically derived lung regions were calculated in EIT images from simulating a homogeneous ventilation distribution in a finite element model. The conventional approach (GI_meas,func ), analyzes images in functionally-defined lung regions, while our proposed measure (GI_meas,anat ) is based on analysis in anatomically-defined regions. We additionally define a simulated comparison (GI_sim,anat ) to determine the lower limit of the GI measure for a homogenous distribution of ventilation.

Results: As expected, the conventional GI_meas,func [0.382 (0.088), median (interquartile range)] were significantly lower than the proposed GI_meas,anat [0.823 (0.152), P<0.05], and were much closer to the lower limit GI_sim,anat [0.343 (0.039)]. Both GI_meas,anat and GI_meas,func were strongly correlated with arterial oxygen partial pressure to fractional inspired oxygen ratio (R=-0.88, P<0.05), whereas GI_sim,anat (R=0.23) was not. GI_meas,anat had a linear-regression slope 3.2 times that of GI_meas,func suggesting a higher sensitivity to the changes in lung condition.

Conclusions: The proposed GI_meas,anat (or shortened as GI_anat ) is an improved measure of ventilation inhomogeneity over GI, and better reflects portion of non-ventilated regions due to alveolar collapse or overdistension.

Keywords: CT segmentation; Global inhomogeneity index; acute respiratory distress syndrome (ARDS); electrical impedance tomography; ventilation heterogeneity.