Abstract
Traditional one-plane (oneP) electrical impedance tomography (EIT) electrode configurations limit evaluation to a convex lens-shaped lung slice, prone to overrepresentation of central areas. Two-plane (twoP) EIT is expected to refine this technique and result in more representative lung imaging.
To compare ventilation distribution using single slice reconstructions, obtained with twoP or oneP.
Computational modelling and in vivo exploratory randomised experimental trial.
A finite element simulation model of oneP and twoP was calculated to estimate captured lung fields. Thereafter, twoP and oneP EIT data were collected in 20 standing horses at baseline, during stance variations and rebreathing. A visual functional region of interest representing the lung field was defined and EIT data reconstructed. Centres of ventilation (CoV), regional ventilation (∆Z, expressed as L or R for left or right lung, and D, CD, CV, or V for dorsal, central dorsal, central ventral or ventral regions) and tidal impedance variation (TIV) were compared between configurations at baseline and between baseline and challenge measurements using Wilcoxon matched pairs signed rank test and a mixed effect model. Proportional changes were compared using a paired t-test.
The simulation revealed a more evenly distributed vertical sensitivity with twoP compared to oneP. In vivo data were analysed to determine which EIT regional parameters increased (twoP vs. oneP). At baseline, ∆ZL
, ∆ZR
and CoV
, were higher when using twoP (7.1 (5.9-8.0) vs. 2.2 (0.9-3.5): p < 0.001; 6.4 (5.6-7.5) vs. 1.5 (0.5-2.79): p < 0.001; 47.7 (46.8-49.29) vs. 43.3 (42.2-45.8): p < 0.001). With rebreathing, an increase in TIV and CoV
was seen with both belt configurations (oneP: 25.13 (±1.967) vs. 49.39 (±11.27): p < 0.001 and 44.3 (±2.300) vs. 45.64 (±2.243): p = 0.002; twoP: 20.21 (±7.398) vs. 49.6 (±13.09): p < 0.001 and 48.79 (±2.126) vs. 50.38 (±1.783): p = 0.005).
TwoP EIT reconstructions resulted in a more vertically uniform representation of ventilation in simulation and in vivo. This likely results in an enhanced representation of peripheral lung fields.