Whole-Head Cocktail Party fNIRS

Whole-Head Cocktail Party fNIRS Whole-head functional near-infrared spectroscopy (fNIRS) recordings from 38 subjects performing covert and overt spatial-attention tasks in a cocktail-party paradigm, with simultaneous eye-tracking. fNIRS: 56 sources, 144 detectors, 1134 measurement channels at ~8.99 Hz. Tasks ----- overt: Spatial attention with overt eye movements: subjects move their eyes to attend to audiovisual stimuli (videos) on the cued side. covert: Covert auditory attention: audio-only stimuli; subjects fixate the central plus sign and attend to left or right without moving their eyes. visualorient: Eye-movement-only baseline. Subjects make ~5 s eye orienting and fixation movements with no audiovisual stimuli and no fixation cross. Matches the overt task structure. resting: Resting-state recording: 5 minutes of central-fixation on a plus sign with no task. longvisualorient: Orienting + visual attention recording with events.tsv aligned to the audiovisual orienting cue. Subjects orient their eyes to the cue; ~15-17 s later a video clip appears on the attended side. Same SNIRF recording as videoattend; only the events.tsv differs. videoattend: Same SNIRF recording as longvisualorient, with events.tsv aligned to the onset of the video clip (~15-17 s after the orienting cue) for video-locked GLM analysis. Privacy ------- This dataset contains only fNIRS optical measurements and eye-tracking time series; no anatomical scans (MRI, CT, photographs, or otherwise) are included. Facial defacing is therefore not applicable to this dataset. Data were acquired and shared under approved IRB consent. Companion analysis code: https://github.com/duwadisudan/wholehead-cocktail-party-fnirs