SNR Optimization – EEG and source-space data of sensory evoked potentials (v1.0)

This dataset supports the SNR_Optimization pipeline for evaluating signal-to-noise ratio (SNR) in ERP source localization. EEG source reconstruction allows estimation of cortical activity from scalp potentials, but the reliability of these estimates depends on experimental and modeling parameters such as trial number, inverse method, and source space resolution. The dataset includes EEG recordings from ten healthy participants collected during electrical stimulation of the right median nerve (2000 trials per participant, 128-channel EEG). It is structured to work directly with the accompanying SNR_Optimization pipeline (GitHub DOI: 10.5281/zenodo.19067281). Dataset Contents subject_1/ – head model used in source reconstruction processed_data/ – participant-specific data folders containing: eeg/ → preprocessed EEG epochs and source spaces per participant leadfield/ → forward model/leadfield matrices per participant current/ → reconstructed source currents and SNR analysis values per participant The folder structure should be preserved in the repository root for proper execution of the pipeline. Preprocessed EEG and current data allow direct generation of figures without re-running initial computations. Usage Notes Run SNR_Optimization.ipynb to reproduce SNR analyses and generate visualizations Dependencies: Python 3.x, MNE library Designed for ERP studies; the dataset can be used to test the impact of trial numbers, source localization methods, and spatial resolution on SNR in both sensor and source space Version and Reproducibility Version: 1.0 Related code: GitHub repository archived in Zenodo (DOI: 10.5281/zenodo.19067281) This dataset enables researchers to evaluate SNR-based convergence of EEG source localization, providing a reproducible and interpretable framework for optimizing experimental and modeling parameters in EEG studies.