Combined external and internal electrodes in EIT for deep changes: A simulation study

Electrical Impedance Tomography can detect changes within human body like tumours, ventilation or cardiac abnormalities using electrodes on the surface. However, EIT images are poor when the changes occur far from the electrodes, in the deep, innermost tissue. A possible method to improve measurements is to position one or more internal electrodes near the expected location of the impedance change when this is known a priori such as during cardiac ablation or other image guided procedures. This paper aims to explore this idea by comparing three different electrode configurations: the conventional external electrodes, the combination of external and internal electrodes and only internal electrodes. The z-score, which expresses the level of distinguishability or sensitivity to an impedance change within the measured object, is used for comparison as it is not biased by the image reconstruction approach. We model a 3D simulated homogenous cylinder tank filled with 0.9% saline containing a small cylindrical anomaly. The internal electrodes were assumed to be located on a cardiac mapping catheter, modeled with electrodes placed on the boundary of a small non-conductive ball, placed near the tank's center. Zero-mean Gaussian noise of 100-500μV was added to simulated electrode voltage measurements. Finally, the results showed that the combination of internal and external electrodes offered the highest consistent distinguishability for most of the patterns of current and voltage measurement and noise levels. The configuration using only internal electrodes starts with worse distinguishability and it decreased dramatically corresponding to the further distance of object.