Single-Subject Motor Imagery EEG With 200 Trials

200 trials of continuous electroencephalography (EEG) recordings from one healthy, right-handed participant performing a left-hand versus right-hand grasping motor imagery task. This single-session subset was compiled by Onyekachi James from the larger WBCIC-MI dataset for educational purposes and was last updated in April 2026. It contains 59 EEG channels and one stimulus channel recorded at 1000 Hz.

Use Cases

• Classify left-hand versus right-hand motor imagery trials using features extracted from the 59 EEG channels and the stimulus trigger channel.
• Benchmark signal preprocessing and filtering techniques on continuous EEG data sampled at 1000 Hz.
• Develop and test feature extraction methods for motor imagery tasks, such as event-related desynchronization/synchronization from the time-series data.
• Train a model to detect the onset of motor imagery events from the continuous recording using the stimulus channel as a ground truth label.

Strengths

• 200 trials provide a balanced sample with 100 trials per class for binary classification.
• High temporal resolution with a 1000 Hz sampling rate across 59 EEG channels.
• Clear provenance as a direct subset of a peer-reviewed dataset published in Scientific Data in 2025.

Limitations

• Data is from a single subject (sub-001), limiting generalizability and population-level analysis.
• Contains only one experimental session, preventing analysis of cross-session or longitudinal variability.

Provenance

Source

Direct subset of the World Robot Conference Contest-BCI Robot Contest Motor Imagery (WBCIC-MI) dataset.

Collection Method

Continuous EEG recordings from a single healthy participant performing a cued motor imagery task.

Time Range

null

Freshness

Dataset subset was last updated in April 2026, though the original source data is from 2025.

Geography

null