In the following repository all codes for reproduce and use the Inner speech Dataset are presented.

The dataset are publicy available at https://openneuro.org/datasets/ds003626

The publicatin are available at https://www.nature.com/articles/s41597-022-01147-2

The stimulation protocol was used for capturing the data, and was developed in Matlab using Psychtoolbox.

The script Stimulation_protocol.m is the main script and uses the other auxiliary functions.

The processing was developed in Python, using mainly the MNE library.

Create an environment with all the necessary libraries for running all the scripts.

conda env create -f environment.yml

Using the Inner_speech_processing.py script, you can easily make your own processing, changing the variables at the top of the script.

The TFR_representation.py generates the Time Frequency Representations used addressing the same processing followed in the paper.

By means of the Plot_TFR_Topomap.py the same images presented in the paper can be addressed.

Please cite this work:

Nieto, N., Peterson, V., Rufiner, H.L. et al. Thinking out loud, an open-access EEG-based BCI dataset for inner speech recognition. Sci Data 9, 52 (2022). https://doi.org/10.1038/s41597-022-01147-2