EEG Recording System and Analysis

EEG recordings were obtained using an 8-channel 24-bit system designed and built by the BioSemi company (Amsterdam, The Netherlands) to our specifications. This “hyper-rate” system is based on their standard systems but with a sampling rate of 262,144 kHz. The very high sampling rate is necessary for the accurate measurement of the electrical stimulation artefacts, which is important when using linear interpolation for removing the stimulation artefacts, as used in the measurement of EASSRs (Gransier et al. 2020), but which is not essential for the ALFIES method. In the BioSemi system, EEG signals are pre-amplified by Ag/AgCI active electrodes placed on the subject’s head. A cap was used to position the recording electrodes according to the international standardized 10/20 system. We placed electrodes at Fz, Fpz, Cz, Iz, P9, P10, and at the left and right mastoids. Except where otherwise stated, analyses are based on the average of the two mastoids and Iz, relative to Cz. Participants sat comfortably in a sound-treated and electromagnetically shielded booth and watched subtitled movies.

EEG recordings were cut into epochs, each starting with a trigger output from the interface with the implant. Because the internal clock of the CI differs from that of the BioSemi system, epochs were further re-cut to have an exact number of electrical pulses (or periods for the analogue dyads) per second. This avoided spectral splatter of the primaries in the subsequent FFT analyses. The epochs were finally averaged into a single 300-ms or 1-s epoch for the analogue and pulse-train stimuli, respectively. The amplitudes and phases at F0, F1, and F2 Hz were then obtained from the corresponding bins of an FFT of that averaged epoch. The power at F0 Hz was compared with that of the adjacent 12 bins (6 each side) using an F test (Dobie and Wilson, 1996). An F ratio greater than 10.92 (p < 0.01, approx. 6 dB signal-to-noise ratio) was deemed significant.