Calculated measures of performance in electrical impedance tomography using a finite-element model

A two-dimensional finite-element model of an electrical impedance tomographic phantom has been investigated. It is perceived that an understanding of the basic measurement system will enable the design of improved data collection hardware, e.g. the number and size of the electrodes. The presented model has sixteen electrodes spaced equidistantly on the perimeter and a centrally placed anomaly. The system is excited using the adjacent-electrode constant-current drive scheme. The size and conductivity of the anomaly are varied in order to investigate how the measurement system will behave to changes in input. Measures of the system's sensitivities to these variations in the anomaly are defined. The four measures derived from the calculated boundary voltages are contrast sensitivity, spatial or radial sensitivity, visibility and rms voltage difference. Results are presented for the four measures, showing how they vary with the parameters of the anomaly. In particular, these results show that there is a clear and distinct difference between spatial and contrast sensitivity. The relationship between the calculated visibility values and the conductivity contrast of the anomaly differs from that predicted theoretically, but is in agreement with the measured relationship. It can also be seen that the presented model is more than twice as sensitive to changes in an anomaly that is more resistive as compared to one that is more conductive than the background or surrounding material.