Descriptor: Dual-Paradigm SSVEP and Motor Imagery 16-Channel EEG Dataset (SMED)

Abstract

This article presents the Steady-State Visual Evoked Potential (SSVEP)-Motor Imagery (MI) Electroencephalography (EEG) Dataset (SMED), a dual-paradigm EEG dataset for brain–computer interface (BCI) research that integrates both paradigms within a single experimental framework. EEG recordings were acquired from 22 participants using a 16-channel OpenBCI system at a sampling rate of 125 Hz under controlled laboratory conditions. The dataset provides synchronized raw EEG signals, event markers, trial labels, and associated metadata for both paradigms, enabling the investigation of exogenous visual and endogenous motor-related neural responses within the same cohort of participants. Unlike conventional MI datasets that focus on unilateral limb imagery, SMED employs whole-body directional MI to capture neural signatures relevant to navigation-style BCI applications. Initial technical validation confirms the presence of frequency-specific SSVEP responses with characteristic harmonic structure, as well as significant task-related neural modulation during MI across central, parietal, and parieto-occipital regions, demonstrating the dataset’s suitability for both paradigm-specific and dual-paradigm research. By combining two representative BCI paradigms in a single resource recorded from the same cohort, SMED can support algorithm benchmarking, cross-paradigm representation learning, transfer learning, and the development of EEG-based BCI systems. IEEE SOCIETY/COUNCIL Systems, Man, and Cybernetics Society (SMCS) DATA TYPE/LOCATION Time-series; Loughborough University, London, U.K. DATA DOI/PID 10.34740/kaggle/dsv/16646295