Errors due to measuring voltage on current-carrying electrodes in electric current computed tomography

It is shown that when a multiplicity of electrodes are attached to a body's surface, the voltage data are most sensitive to changes in resistivity in the body's interior when voltages are measured from all electrodes, including those carrying current. This assertion, which is true despite the presence of significant levels of skin impedance at the electrodes, is supported both theoretically and by experiment. Data were first taken for current and voltage at all electrodes. Then current was applied only at a pair of electrodes, with voltages measured on all other electrodes. Targets could be detected with better signal-to-noise ratio by using the reconstructed data than by using the directly measured voltages on noncurrent-carrying electrodes. Images made from voltage data using only non-current-carrying electrodes had higher noise levels. It is concluded that in multiple electrode systems for electric-current-computed tomography, current should be applied and voltage should be measured from all available electrodes.