Intentionally vs. Spontaneously Prolonged Gaze: A MEG Study of Active Gaze-Based Interaction

This dataset accompanies the published paper "Intentionally vs. Spontaneously Prolonged Gaze: A MEG Study of Active Gaze‑Based Interaction" (Cortex, 189:76–96, 10.1016/j.cortex.2025.05.010; Vasilyev et al., 2025), which explores neural mechanisms underlying intentional and spontaneous gaze behavior during gaze‑based human–computer interaction. Magnetoencephalography (MEG) and eye‑tracking data were collected from participants engaged in a specially designed gaze‑controlled computer game, where actions required intentional fixation of visual targets. Sample Size Participants: 29 healthy individuals (mean age: 25 years; 12 females). Trials: 6,748 total trials, with each participant contributing over 100 trials per condition. The dataset is balanced, containing an equal number of trials for the Intentional and Spontaneous conditions. Recording Setup MEG: 306-channel Elekta Neuromag TRIUX™ system. Eye-tracking: EyeLink 1000 Plus (monocular recording, 1000 Hz). Stimuli and Data Labeling Participants played the gaze-controlled game "EyeLines". Each move required selecting an object (a colored ball) by maintaining gaze fixation on it for at least 500 ms, after which a visual "selected" feedback animation appeared. To complete the selection, participants had to subsequently perform a confirmation dwell on a designated visual "button". Gaze dwells on balls occurred in two distinct contexts: Intentional dwell: Participants purposefully fixated on a ball intending to select it. This dwell was followed by a saccade toward the confirmation button approximately 700–800 ms after initial fixation onset. This interval includes: Initial dwell: 500 ms Feedback appearance: approximately 60–100 ms Feedback processing and saccadic response initiation: approximately 100–150 ms Spontaneous dwell: Participants fixated a ball spontaneously during gameplay, typically while deliberating potential moves, analyzing ball positions, or visually exploring the game board without explicit intention to select. These spontaneous dwells also triggered the "selected" animation but were not followed by confirmatory button dwell. Epoch Window MEG data epochs represent signals related to each dwell, spanning from −2000 ms to +1999 ms relative to dwell-initiating fixation onset. Due to the game design, controlled comparisons between Intentional and Spontaneous epochs are valid only within the 0 to 500 ms interval, as behavioral activity following feedback varies significantly between conditions. Data Preprocessing MEG data underwent initial post-acquisition preprocessing using the MaxFilter routine, including: Temporal signal space separation (tSSS) with 10 s buffer and correlation threshold of 0.9 Movement correction (MaxMove) Head-origin normalization This preprocessing is part of post-acquisition handling and represents the raw data state for this dataset. Epoch synchronization between MEG and eye-tracking data was ensured by matching synchronization triggers generated by the eye-tracker server with MEG triggers.