Multifrequency image reconstruction for electrical impedance tomography

Abstract Electrical Impedance Tomography (EIT) is a medical imaging technique that is primarily used to monitor the respiration of a patient. Because EIT is based on electrical measurements, it is a safe, non-invasive, and cost-effective imaging technique. However, the EIT image reconstruction is a severely ill-posed problem that gives low spatial resolution where only large variations in tissue conductivity can be visualized. Furthermore, widely used time difference EIT relies on a single frequency alternating current measurement which does not allow for discrimination of different tissues on their conductivity spectra. Here we show the application of a new EIT reconstruction algorithm which correlates measurements taken at different frequencies to include the spectral dependency of the tissue properties. The algorithm is tested on a simulated phantom using data for muscle and lung tissue from the literature. It shows that contrary to a standard EIT image reconstruction, the frequency dependence of the tissues is retained, which can be used to further improve distinguishability in EIT images.