MET - Music-Induced Emotion EEG Dataset

# MET: Music-Induced EEG Dataset ## Overview The MET (Music-Induced EEG Dataset) contains EEG recordings collected from 20 participants during music-listening experiments. Each participant completed: - Twenty music-listening sessions are named as ses-txx. (eg. ses-t01 means the song index is 1) Each music session corresponds to a different song. The total recording duration for each song is 130 seconds, including: - 5 seconds before music onset - 120 seconds of music playback - 5 seconds after music offset EEG signals were originally recorded using a 128-channel EEG system at a sampling rate of 1000 Hz. After preprocessing and channel selection, 66 EEG channels are provided in this dataset. The dataset follows the Brain Imaging Data Structure (BIDS) specification for EEG recordings. --- ## Repository Structure participants.tsv: subject information and emotion labels (valence & arousal level) from the questionnaire they answered. Each EEG session contains the following files: *_eeg.eeg: EEG signal data in BrainVision format. (Unit: µV [microvolts]) *_eeg.vhdr: BrainVision header file containing acquisition parameters. *_eeg.vmrk: Event marker file. *_eeg.json: EEG recording metadata. *_channels.tsv: Information about EEG channels, including channel names and units. *_events.tsv: Timing information for experimental events. *_events.json: Description of event variables stored in events.tsv. *_electrodes.tsv: Spatial coordinates of EEG electrodes. *_coordsystem.json: Coordinate system used for electrode locations. *_scans.tsv: List of files acquired during the session. --- ### Dataset Usage The dataset contains: - EEG recordings from 20 participants - One resting-state session per participant - Twenty music-listening sessions per participant - Subjective ratings including: - Valence (1–9) - Arousal (1–9) - Likeability (1–5) - Familiarity (1–5) - Song-level affective labels: - HVHA (High Valence High Arousal) - HVLA (High Valence Low Arousal) - LVHA (Low Valence High Arousal) - LVLA (Low Valence Low Arousal) - N (Neutral) Users may utilize either the provided affective categories or the subjective ratings as prediction targets. ### Recommended Tasks Possible benchmark tasks include: 1. Emotion category classification (HVHA, HVLA, LVHA, LVLA, N) 2. Valence prediction 3. Arousal prediction 4. Likeability prediction 5. Familiarity prediction 6. Cross-subject EEG classification 7. Music-evoked neural pattern analysis --- ## Sample Code (PYTHON): ```py from mne_bids import BIDSPath, read_raw_bids # 1. Set file path and file name bids_path = BIDSPath( subject='01', # subject index (eg. 01) session='t01', # trial index (eg. t01) task='music', datatype='eeg', root='/Users/kathy/Documents/EEG data/EGI_preprocessed_bids' # directory root path ) # 2. Load BIDS file raw_read = read_raw_bids(bids_path=bids_path, extra_params={'preload': True}) # 3. Check Info print(raw_read.info) ``` --- References ---------- Appelhoff, S., Sanderson, M., Brooks, T., Vliet, M., Quentin, R., Holdgraf, C., Chaumon, M., Mikulan, E., Tavabi, K., Höchenberger, R., Welke, D., Brunner, C., Rockhill, A., Larson, E., Gramfort, A. and Jas, M. (2019). MNE-BIDS: Organizing electrophysiological data into the BIDS format and facilitating their analysis. Journal of Open Source Software 4: (1896). https://doi.org/10.21105/joss.01896 Pernet, C. R., Appelhoff, S., Gorgolewski, K. J., Flandin, G., Phillips, C., Delorme, A., Oostenveld, R. (2019). EEG-BIDS, an extension to the brain imaging data structure for electroencephalography. Scientific Data, 6, 103. https://doi.org/10.1038/s41597-019-0104-8