Validation of Neuphony 8-channel EEG flex cap: a comparative study with BioSemi 64-channel EEG system.

Background: Electroencephalography (EEG) is a widely used tool in neurological assessment and cognitive research due to its reliability, high temporal resolution, and non-invasive nature. While portable EEG systems offer improved accessibility, many rely on electrode caps or extensive electrode arrays, which limit ease of use and portability in applied settings. The present study aimed to evaluate whether the 8-channel wearable Neuphony EEG flex cap can reliably capture key event-related potential (ERP) components such as P300 and mismatch negativity (MMN), while also providing robust measures of continuous EEG activity. Methods: This study involved 25 healthy participants, with an average age of 19.72 ± 0.8 years. EEG recordings were obtained using the Neuphony Flex Cap and BioSemi system while subjects engaged in the following tasks: resting state, auditory oddball, and visual discrimination tasks. Participants were instructed to keep their eyes open and concentrate on a cross during the resting state recording. Two distinct sound tones were played during the auditory oddball task to elicit P300 and MMN responses. For the visual discrimination task, participants were asked to respond upon seeing either an M or a W among distractors. Accuracy and reaction times were recorded for both tasks. Results: We evaluated the average relative PSD of the Neuphony EEG flex cap and BioSemi Active Two system across the entire frequency range (1-30 Hz), revealing a significant difference in the 8 to 13 Hz band. A repeated measures ANOVA showed a significant main effect of device for P300 amplitude [F(1, 24) = 271.59, p < 0.001, ηp2 = 0.919] and P300 latency [F(1, 24) = 8298.6, p < 0.001, partial η2 = 0.997] reflecting overall differences in recording between devices. Similarly for Late Positive Complex (LPC), a repeated measure ANOVA showed a significant main effect of device for amplitude [F(1, 24) = 16.44, p = 0.00046, partial η2 = 0.406] and latency [F(1, 24) = 7094.90, p < 0.001, partial η2 = 0.997]. Discussion: This research indicated that it is feasible to obtain fair-quality EEG data for research purposes using the Neuphony EEG flex cap. However, the strength of the frequency band between 8 and 13 Hz showed variability, as did the event-related peaks. Event-related potentials (ERPs) were comparable for both devices, with a slight consistent delay in the 8-channel Neuphony EEG flex cap, potentially due to Bluetooth signal transmission.