Tactile-evoked EEG Dataset for Natural Perception Using an Integrated Stimulation-Recording Framework

The increasing demand for assistive living and medical technologies in aging societies has driven advancements in tactile-evoked Brian Computer Interface (BCI) systems, offering an alternative to traditional visual and auditory-based BCI systems. However, the development of such systems is constrained by challenges in quantifying tactile sensations and a lack of diverse datasets. This study presents an integrated system enabling natural tactile perception during dynamic touch experience while simultaneously recording electroencephalographic (EEG) responses. EEG signals were collected from 10 healthy participants (64 channels, 1000 Hz) in natural tactile perception tasks involving contact with three distinct materials. Preliminary analysis revealed significant differences in the P300 peak latency and amplitude between tactile conditions, highlighting the unique characteristics of tactile-evoked EEG signals. A three-class classification using Common Spatial Pattern (CSP) and Support Vector Machine (SVM) models demonstrated above-chance accuracy. This tactile-evoked EEG dataset provides a valuable resource for seeking tactile-related neural mechanisms and driving the practical application of BCI systems, offering a pathway to improved user experiences and functionality in real-world scenarios.