Subsecond observations of EIT voltage changes on the human scalp due to brain stimulus

A pilot study has investigated the feasibility of imaging human brain function using an electrical impedance tomography (EIT) system time-locked to an evoked response (ER) system. A sixteen-electrode planar EIT configuration was used with polar current injection. We report here measurements on two volunteer patients who were fully awake in all tests. Reference data (with no applied stimulus) yield EIT nearest-neighbour voltage differences over the range 5-28 mV. In comparison with forward calculations [C.M. Towers et al. (2000)], these data suggest a value of skull conductivity of the order of 0.05 S/m. Visual and auditory stimuli were applied as single discrete events to evoke neural responses (VER and AER respectively). In each case, EIT data acquisition commenced at a time between 70 and 740 ms later, taking 308 ms to complete. Average values of voltage pair data over many frames are presented here, with the emphasis on the VER data. When comparing data taken under stimulus conditions against the reference data, voltage differences of up to approximately 3 mV are observed in both AER and VER cases. We attribute these voltage changes to synaptic activity. Preliminary reconstructed images of conductivity are discussed.