EEG dataset on auditory streaming and interoception, collected at the University of Chicago and organized in BIDS format. The README gives few details of the experimental design, so the description rests mainly on the dataset name and metadata.
This dataset contains 1,983 overnight polysomnography recordings from patients at the Massachusetts General Hospital sleep laboratory, most of whom were evaluated for suspected sleep disorders such as obstructive sleep apnea. It was created for the PhysioNet/Computing in Cardiology Challenge 2018 on automatic detection of sleep arousals. The training set (994 subjects) includes expert annotations of sleep stages, respiratory events and arousals, while the test set (989 subjects) contains signals only. Recordings comprise EEG, EOG, EMG, respiratory effort, airflow, SpO2 and ECG channels sampled at 200 Hz.
EEG dataset on auditory streaming and interoception, collected at the University of Chicago by Veillette, Joshi, Li, and Nusbaum. Data are organized in BIDS format (BrainVision files) and released under a CC0 license.
EmoEEG-MC is a multi-context emotional EEG dataset containing 64-channel EEG and peripheral physiological signals (PPG and GSR) from 60 participants. Emotions were induced in two contexts, video viewing and mental imagery, and covered seven categories: joy, inspiration, tenderness, fear, disgust, sadness and neutral. The dataset is intended to support cross-context emotion decoding, and SVM baselines with L1 regularization achieved above-chance accuracy for binary and seven-category classification.
Intracranial electrophysiology recordings from the ventral posterolateral (VPL) thalamus of macaques, collected with a 32-channel linear probe during transcranial focused ultrasound (FUS) neuromodulation and tactile stimulation. FUS was targeted either to VPL thalamus or to insular cortex while recording remained in VPL. The dataset includes spiking and local field potential recordings supporting a study of FUS effects on primate thalamic activity.
EEG and behavioral dataset from a prolonged simulated driving study in which participants completed two 45-minute driving sessions with lateral perturbations and auditory cues. Cues were either random or predictive of perturbation onset. The study was designed to identify driver fatigue biomarkers from EEG and to compare them with objective driving performance measures. Data are in BIDS format and include EEG, eye tracking and vehicle logs.
A raw multimodal dataset from 30 healthy right-handed adults performing a hierarchy of cognitive, motor, and combined cognitive-motor tasks across three sessions. It includes EEG, fNIRS, ECG, EMG, Biodex torque/kinematics, push-button responses, and subjective ratings of sleepiness and cognitive load. The data are organized in BIDS and are unprocessed exports from the acquisition systems.
EEG, galvanic skin response (GSR) and electrocardiogram (ECG) recordings from healthy adult volunteers listening to 40-s synthetic music clips designed to target specific affective states defined by valence and arousal. The data were collected to calibrate an affective brain-computer music interface that induces target emotional states through real-time modification of generated music. Each session comprised 5 runs of 18 music trials, and self-reported valence and arousal ratings are included. This is the calibration dataset of three related datasets from the BCMI-MIdAS project.
EEG recordings from 25 participants performing a two-competing-speaker selective attention task, in which they attended to either the left or right audio book. The participant's own name was embedded 10 times per block in the to-be-ignored stream, allowing study of attention capture and neural tracking of attended and ignored speech. Blocks alternated between an omnidirectional condition (equal loudness) and a beamforming condition (attended stream louder).
EEG recordings from 142 participants who received laser-evoked pain stimuli at fixed low and high intensities and rated the pain intensity after each stimulus. Each participant completed 3 blocks of 10 stimuli, with high and low pain stimuli evenly distributed within each block. Data were recorded with a Biosemi system and are intended for studying the neural correlates of pain perception.