Frequency-difference electrical impedance tomography: Phantom imaging experiments

Abstract

Frequency-difference electrical impedance tomography (fdEIT) using a weighted voltage difference has been proposed as a means to provide images of admittivity changes at different frequencies. This weighted difference method is an effective way to extract anomaly information while eliminating background effects by unknown boundary geometry, uncertainty in electrode positions and other systematic measurement artefacts. It also properly handles the interplay between conductivity and permittivity in measured boundary voltage data. Though the proposed fdEIT algorithm is promising for applications such as detection of hemorrhagic stroke and breast cancer, more validation studies are needed. In this paper, we performed two-and three-dimensional numerical simulations and phantom experiments. Backgrounds of imaging objects were either saline or carrot pieces suspended in saline. We used carrot pieces to simulate a more realistic frequency-dependent admittivity distribution. Test objects were banana, potato or conductive gel with known admittivity spectra. When the background was saline, both simple and weighted difference approaches produced reasonably accurate images. The weighted difference method yielded better images from two-dimensional imaging objects with background of carrot pieces. For the three-dimensional head-shaped phantom, the advantage of the weighted frequency difference method over the simple difference method is not as obvious as in the case of the two-dimensional phantom. It is unclear if this is due to measurement errors or limitations in the linear algorithm. Further refinement and validation of the frequency difference image reconstructions are currently in progress.