Feasibility of dry surface electrodes for electroencephalography (EEG) in dogs

Introduction: Epilepsy is an extremely common neurological problem affecting dogs. One significant challenge is determining whether a dog’s paroxysmal episode is the result of seizure activity. Electroencephalography (EEG) is the only diagnostic tool that confirms seizure activity. Despite this, EEG is underutilized in veterinary medicine likely due to difficulties in equipment setup, e.g., painstaking subcutaneous electrode placement or sedation for instrumentation. Novel dry surface electrode systems investigated for people are less invasive than current minimally invasive subdermal wire electrodes (SWE) and may hasten EEG instrumentation in dogs. Objectives: The aim was to evaluate the clinical feasibility of dry surface electrodes for long-term EEG recording in dogs. We hypothesized that the dry surface electrodes would remain functional and perform equivalently to the SWE. Materials and Methods: One six-hour EEG was recorded from each of six awake healthy dogs. Four PressOn and four Spring-loaded dry surface electrodes were randomized each to one of eight locations above a gold standard SWE, creating four comparison pairs. Outcome analysis examined both quantitative and qualitative variables: power spectrum, electrode attachment, and artifact. Results: Power spectrum analysis showed good agreement between both dry electrodes and SWE for all frequency bands, but with a significant statistical bias - the dry electrodes consistently read higher. Impedance was significantly higher in both dry electrodes compared with the SWE. Overall, SWE electrodes were 3.14 times more likely than PressOn and 43.36 times more likely than Spring-loaded electrodes to remain attached. Movement artifact was significantly higher in both dry electrodes compared with SWE. Muscle artifact was significantly higher in SWE compared with both dry electrodes. The proportion of skin irritation was significantly higher in both dry electrodes compared with SWE. Conclusions: Both dry electrodes demonstrated a good power spectrum agreement with the SWE, indicating a clinically reliable signal. However, for most of the other outcome variables assessed, SWE consistently outperformed the novel dry electrodes. The PressOn electrode have potential for clinical use given the relatively good attachment properties. Both dry electrodes had unacceptably high impedances. This weakness will need to be overcome prior to their use in a clinical setting.