Computational protocol: Discrete and continuous mechanisms of temporal selection in rapid visual streams

Similar protocols

Protocol publication

[…] Brain signals were continuously recorded with a 306-channel whole-head magnetometer (Elekta Neuromag®, Sampling rate: 1000 Hz; High pass filter: 0.1 Hz; Low pass filter: 330 Hz) within a room shielded against electromagnetic noise (Maxshield). MEG channels were organized in 102 triplets composed of one magnetometer and two orthogonal planar gradiometers. In addition, electrocardiogram, and vertical and horizontal electro-oculograms were also recorded for offline rejection of artefacts induced by eye movements and heartbeat. Subjects’ head positions were tracked with four coils placed over frontal and mastoïdian skull areas and measured at the beginning of each run with an isotrack polhemus Inc. system. Head positions were then realigned on the position of the first run in order to compensate for head movements between runs. Signal Space Separation was applied to MEG signals in order to decrease the impact of magnetic sources outside the sensor helmet. Magnetometers and gradiometers were visually inspected to identify bad channels (1–7 bad channels across subjects). Head movement compensation, bad channel correction and signal space separation were applied using MaxFilter software (Elekta®).Continuous data were then epoched with the Fieldtrip software (http://www.fieldtriptoolbox.org/). Localizer epochs range from −200 to 600 ms after stimulus presentation. Dual-task epochs started 500 ms before T1 onset and ended 2000 ms later during. A baseline correction was applied for each trial and each sensor using the time period before stimulus onset. A panel of measures (variance, minimum, maximum, range) were then computed across sensors and displayed in scatter plots in order to identify and reject the trials that might be artifacted (mean proportion of rejected trials per subject: M = 4.68%, SD = 3.77). Independent component analyses were applied separately for each type of sensor using fastICA algorithm. Components topographies were visually inspected and their time courses were correlated with the EOG and ECG signals. The components related to the cardiac artefact or to the eye movements were then rejected from the raw data. [...] An anatomical MRI (3T Siemens MRI scanner with a spatial resolution of 1 × 1 × 1.1 mm3) was acquired for each subject. Subjects’ head were digitized and their position inside the sensor helmet was tracked in order to coregister the MEG signals and subjects’ anatomy. Gray and white matters were segmented with the Freesurfer software, (http://surfer.nmr.mgh.harvard.edu/). Cortical surfaces were reconstructed with Brainstorm©. Models of the cortex and of the head were used to estimate the current-source density over the cortical surface. The forward model was computed with overlapping spheres analytical model. Weighted minimum norm estimate (wMNE) was used for inverse modeling (depth-weighting factor: 0.5) and dipole orientations were constrained to be normal to the cortical mantle. In order to perform group analyses, the source estimate data of each individual were projected on the freesurfer standard anatomical template (an average brain based on 40 subjects). Single subject MEG signals were transformed in Z-scores relative to baseline and spatially smoothed over 10 mm.MEG, MRI data, and analyses code are available upon request. […]

Pipeline specifications

Software tools FieldTrip, fastICA, FreeSurfer, Brainstorm
Applications Miscellaneous, Magnetic resonance imaging
Organisms Homo sapiens