3D EIT Improves Detection of Ventilation Inhomogeneity in Obstructive Lung Disease

Abstract: Obstructive lung diseases are highly heterogeneous. Using measurements of equine asthma, we compared 2D and 3D EIT metrics against lung mechanics. 3D EIT showed greater sensitivity to treatment effects, indicating improved detection of ventilation heterogeneity. Introduction: Traditional 2D EIT uses a single electrode plane, reconstructing a transverse slice, but obstructive lung diseases (OLD) exhibit heterogeneity along the craniocaudal axis, potentially missed by 2D imaging. 3D EIT, using multiple electrode planes, provides volumetric reconstructions, capturing a larger fragment of lung geometry. We compare 2D and 3D EIT in assessing ventilation inhomogeneity in OLD, using an equine asthma model, and hypothesize that 3D EIT offers superior sensitivity. Materials & Methods: Data were acquired from ten horses with severe equine asthma during exacerbation and remission phases, as described in [1]. EIT images were reconstructed in 2D- and 3D (example Figure 1) and EIT metrics computed: Global Inhomogeneity (GI) [2], Regional Ventilation Delay (RVD) standard deviation [3], and Global Asynchrony Index (GAI) from Pixel Asynchrony Value (PAV) [4]. All metrics were calculated for 2D and 3D ROIs defined by 20% ventilation threshold. Paired t-tests compared metric changes against lung resistance RL reductions. Results: Paired comparisons (Exacerbated-Treated, n=10; Before-After salbutamol, n=9) showed RL decreases (p < 0.001). Paired t-test results for 2D and 3D EIT metrics (Table 1) show 3D metrics consistently have higher significance than 2D, especially for GAI and GI, while 2D SD-RVD failed to detect consistent changes. Figure 2 presents the paired 2D/3D comparisons for GAI, SD-RVD and GI. Conclusions: The 3D reconstruction better captures craniocaudal ventilation variations and aligns more closely with lung mechanics. We thus recommend 3D EIT for imaging of obstructive lung diseases.