When preprocessing the BCICIV 2a motor imagery (MI) dataset, we noticed a discrepancy between the theoretical total number of trials and the valid samples used in the paper:
- Theoretical total trials: 9 subjects × 2 sessions × 288 trials = 5184
- Valid samples in the paper: 5088 (a difference of 96 trials)
We would greatly appreciate it if you could clarify:
Whether the 96 missing trials are invalid trials officially excluded from the BCICIV 2a dataset (e.g., due to subject inattention, excessive EEG artifacts, equipment malfunctions during data collection)?
Are there any specific preprocessing exclusion criteria (e.g., amplitude thresholds, signal-to-noise ratio) that lead to the reduction from 5184 to 5088 valid samples?
This information is critical for us to align our data preprocessing pipeline with the settings described in the paper. Thank you for your time and response!
When preprocessing the BCICIV 2a motor imagery (MI) dataset, we noticed a discrepancy between the theoretical total number of trials and the valid samples used in the paper:
We would greatly appreciate it if you could clarify:
Whether the 96 missing trials are invalid trials officially excluded from the BCICIV 2a dataset (e.g., due to subject inattention, excessive EEG artifacts, equipment malfunctions during data collection)?
Are there any specific preprocessing exclusion criteria (e.g., amplitude thresholds, signal-to-noise ratio) that lead to the reduction from 5184 to 5088 valid samples?
This information is critical for us to align our data preprocessing pipeline with the settings described in the paper. Thank you for your time and response!