Quantification improvement in EIT reconstruction algorithms by incorporating a priori geometrical information

In this paper we develop a method that introduces a priori geometrical information about a conductivity distribution in EIT reconstruction algorithms in order to improve their quantification ability. We propose correction expressions for two reconstruction algorithms: backprojection algorithm and exponential algorithm. There is no theoretical evidence of the uniqueness of the solution to the inverse boundary problem which EIT deals with. We focused our attention to the problem of a circular centred uniform perturbation in a uniform background. It is easily seen that for cosine profiles there are different combinations of conductivity change and perturbation radius that yield the same voltage change. A similar result is obtained for adjacent injection profiles. This supports the fact that the kind of correction we propose is necessary not only as a consequence of the behaviour of the reconstruction algorithm but also because of the nature of the problem