The ability to image the impedance changes that accompany neuronal depolarization in the brain would constitute a major advance in neuroscience technology. Unfortunately, these changes are likely to be small and rapid and so difficult to measure. The impedance change at frequencies above 10 kHz, as used by conventional EIT systems, may be estimated to be about 0.1%. Modelling indicates that a much larger impedance change of about 7% may occur with DC or very-low-frequency excitation. Difficulties with this approach include a low permissive current level and high electrode impedance. We constructed a prototype system employing square wave excitation at 5 Hz to evaluate such problems. It was tested in a saline-filled tank, recording 4000 frames s-1 at a current level of 50 microA. After averaging 100 sets of frames, the signal to noise ratio was 40-50 dB, and reciprocity errors were mostly 10-20%. Images of discrete resistivity changes of less than 10% could be obtained, but with significant systematic errors. While our prototype would not be suitable for neurophysiological imaging as it stands, it has enabled us to determine the modifications that would be required to construct a system for this application.