Objective: Tissues in the body differ by their frequency-dependent conductivity. Frequency-differential electrical impedance tomography (fdEIT) is a promising technique to reconstruct the distribution of tissue inside the body by injecting current at two frequencies and measuring the resulting surface-potential.
Approach: The Gauss-Newton method is one way to map the surface measurements to a conductivity image. Usually, the minimization function contains only weighted differential measurement data and a regularization. This traditional method is extended by absolute measurement data to improve fdEIT reconstruction results. The key challenge of unknown torso geometries and electrode displacement has been addressed for the reconstruction of different lung pathologies.
Main results: The frequency-dependent conductivity of the background was reconstructed precisely and a contrast between organs was achieved. The algorithm shows good performance compared to GREIT and the traditional Gauss-Newton method with respect to the figures of merit of GREIT.
Significance: The reconstruction is robust in the presence of noise. One application of the algorithm might be the detection and monitoring of lung diseases like edema or atelectasis.