A Modified Noise Model of Electrical Impedance Tomography System by Considering Colored Noises

Sufficient signal-to-noise ratio (SNR) is critical to guarantee the satisfactory performance of an electrical impedance tomography (EIT) system. In this article, a thorough SNR analysis of EIT systems is presented. Besides the well-known random noise, the influences of colored noises, such as switching noise, low-frequency noise, and harmonic noise, are considered. By investigating the colored noises, a novel noise model of electrical impedance tomography systems is first proposed, which is validated by both simulation and experiment conducted in a laboratory-built electrical impedance tomography system. Finally, different conventional methods to reduce the colored noises are discussed and associated circuit-level considerations are evaluated by applying a front-end isolation structure in the laboratory-built EIT system. Overall, this work provides a more practical noise model of electrical impedance tomography systems by including the colored noises. It can serve as a guide to improve the SNR and therefore the performance of an electrical impedance tomography system for high-demanding biomedical applications. © 1963-2012 IEEE.