An EEG Imagined Speech Dataset for BCI Applications with Binary and Numerical Stimuli

The dataset contains electroencephalographic (EEG) recordings were collected from 30 healthy university students (mean age 20.53 ± 1.54 years; 8 female, 22 male) during imagined speech tasks for Brain-Computer Interface (BCI) applications. All participants provided written informed consent, and the study was approved by the Institutional Research Ethics Committee of Tecnológico de Monterrey (protocol code CA-EIC-035, July 2024).Utilizing the Unicorn Hybrid Black wireless headset (g.tec neurotechnology GmbH) with 8 electrodes which were ccommodated according to the 10-20 international system (Fz, C3, Cz, C4, Pz, PO7, Oz, PO8) with a sampling rate of 250 Hz at 24-bit resolution. Two paradigms were implemented in Spanish: a binary paradigm ('Yes' and 'No') and numerical paradigm ('One', 'Two' and 'Three').This dataset is realeas under CC BY-NC-SA 4.0 license and is intended to support further exploration in computational neuroscience, biomedical engineering, and artificial intelligence.