Using variational regularization method of electrical impedance tomography to reduce the position error of human arm

This paper investigates how to reconstruct human arm images using electrical impedance tomography. A simplified arm model is first established. The comparison between the Tikhonov regularized reconstruction algorithm and the Variation regularized reconstruction algorithm is then made. The results show that the Variation regularized reconstruction algorithm has the lesser position error on the organizational junction edges and is basically overlap with the target area of the model. How to eliminate the artefact of the reconstructed image needs to be further investigated. (paper)