Improvement in EIT image reconstruction by using double FEM mesh

We propose the use or two separate meshes; one (line with 1408 elements) to model the potentials more accurately in the forward problem, the other (coarse with 88 elements) to model the inverse problem conductivity distribution. Static images are reconstructed using Gaussian noise added Ankara standard data set with the Newton-Raphson algorithm. The quality of the images is evaluated in terms of spatial resolution calculated as the full-width-half-maximum, for single and double, conducting and insulating objects at different positions along the diameter. Results show that better images are reconstructed with lower error end an Increase in resolution does appear to be achieved. Although the method proposed has been used in other arena of imaging, it is difficult to find an explicit description of this method in the literature so in this sense the suggestion is novel.