Laboratory Scale Tests Of Electrical Impedance Tomography

Electrical impedance tomographs (magnitude and phase) of known, laboratory-scale<br>targets are reported. Three methods are used to invert electrical impedance data and their<br>tomographs compared. The first method uses an electrical resistance tomography (ERT)<br>algorithm (designed for DC resistivity inversion) to perform impedance magnitude<br>inversion and a linearized perturbation approach (PA) to invert the imaginary part. The<br>second approximate method compares ERT magnitude inversions at two frequencies and<br>uses the frequency effect (FE) to compute phase tomographs. The third approach,<br>electrical impedance tomography (EIT), employs fully complex algebra to account for<br>the real and imaginary components of electrical impedance data. The EIT approach<br>provided useful magnitude and phase images for the frequency range of 0.0625 to 64 Hz;<br>images for higher frequencies were not reliable. Comparisons of the ERT and EIT<br>magnitude images show that both methods provided equivalent results for the water<br>blank, copper rod and PVC rod targets. The EIT magnitude images showed better spatial<br>resolution for a sand-lead mixture target. Phase images located anomalies of both high<br>and low contrast IP and provided better spatial resolution than the magnitude images.<br>When IP was absent from the data, the EIT algorithm reconstructed phase values<br>consistent with the data noise levels.