Optimized a priori D-bar Reconstructions of Experimental 2-D EIT Data
Recent advances in D-bar methods for Electrical Impedance Tomography have resulted in a priori techniques, in which prior information is included directly in the equations for D-bar. If properly performed, these techniques lead to improved spatial resolution in the final reconstructions. One of the main challenges in these a priori methods has been how to select a priori estimates for conductivity values, since poor guesses can result in artifacts and distortions in the resulting reconstructions. These problems arise in part due to a mismatch between the scattering transforms computed from the measured data vs. the a priori data. In this talk, we present a novel method for choosing conductivity values in the prior, wherein an optimization routine is used so as to minimize the mismatch in the scattering data. The new algorithm is validated using reconstructions of experimental tank data and experimental human data.