Computational protocol: Short-term EEG dynamics and neural generators evoked by navigational images

Similar protocols

Protocol publication

[…] The EEG was recorded from 128 scalp sites using a shielded electrocap. All recordings were referenced to the left earlobe electrode. Vertical and horizontal eye movements (EOG) were recorded bipolarly. All electrode impedances were maintained below 5 kΩ. Scalp potentials were amplified by ANT DC-amplifiers (ANT, the Netherlands) and digitized with a rate of 2.048 Hz, a resolution of 16 bit (range 11 mV). A band pass from DC to 256 Hz and a notch filter (47.5–52.5 Hz) were also applied. Participants were asked to avoid eye blinks and to fixate the green dot presented in the middle of the screen in order to reduce eye artifacts. In order to verify the effectiveness of the eye fixation requirement, the number of eye movements was recorded throughout the different visual stimulation periods. For this, saccades including small saccades of about 0.8° and other eye movements were automatically selected by a Matlab (MathWorks Inc) script using eye velocity threshold []. This selection was then verified by visual inspection. For all subjects, the fixation requirement was respected. Only 0.19 ± 0.09 saccades per second were recorded and this mean number was not different depending on the different images.Off-line treatment and statistics were performed by means of EEGLAB software [] and artifactual portions of the EEG data were rejected after appropriate independent component analysis (ICA). [...] The statistical significance of difference between the two experimental conditions (“checkerboard” and “3D-tunnel”) amplitude and respective latency in time (ERP) and in time-frequency domain analysis (ERSP and ITC) was calculated with ANOVA (and post-hoc Freeman correction) in the selected EEG channels. For each stimulation, the peak latency and mean amplitude around the peak (ranging from -20 ms to 20 ms) were measured [] for the components P100, P200 in posterior areas and concomitant N100, N200 in frontal areas and late centro-parietal P300. In order to estimate the effect size (d) on the statistical significance, post hoc power analyses were performed with the G*Power 3.1 software []. The power (1-β) was computed for all the significant P < 0.05. For significance in the full scalp array in the ERP, ERSP and ITC, we employed a nonparametric permutation and the Holm’s method [] to correct for multiple comparisons. This method is provided by EEGLAB software [].To find the generators of the P100, P200 and P300 components, a rigorous method was used to establish a threshold value that helps in identifying statistical significance of the current density magnitude. To do so, we used the nonparametric permutation test described in Nichols and Holmes []. The rationale for using this method for the inverse solution was explained in details in []. In order to use the t test as the value of merit, we subtracted one to each voxel. In order to build the empirical distribution for the Holmes method, we randomly multiplied the value of each voxel by -1 or 1. Finally, we performed a total of 8192 t-tests for each component analyzed. The normalization process is described by the following formula: Ji(t)normalized=Ji(t)∑voxel=0nVoxel=2030Jvoxel(t)−1 where Ji(t) represents the inverse solution for the i-th voxel for the time t. […]

Pipeline specifications

Software tools EEGLAB, G*Power
Applications Miscellaneous, Clinical electrophysiology