Influence of Boundary Deformation on Image Reconstruction in Electrical Impedance Tomography

For electrical impedance tomography (EIT) lung imaging, chest expansion and contraction could change the electrode placement and deform the boundary of the image area, which changes the electric field distribution and distorts the reconstructed images. EIT images of lung function often produce severe artifacts due to the expansion of the thoracic cage, so it is necessary to study the effect of thoracic boundary deformation on EIT images. In this paper, we established several elliptical image reconstruction models, which deform from 1% to 15% in the anterior–posterior (AP) dimension from the circle field domain, to emulate chest expansion. The Quantitative evaluation of the influence of the deformation on the reconstruction images has been made. Images are estimated by parameter DORI and SSIM. The results show that these two parameters can evaluate image quality effectively, and image error increases with the increasing of boundary deformation. Although the error is small when the deformation extent of the boundary shape is small, images are seriously distorted when the deformation increases. Image error increases significantly when the boundary deformation increases, which may lead to image artifacts and maybe result in misdiagnosis.