Materials

Dataset

This project used EEG recordings of a feasibility clinical trial that tested a P300-based Brain Computer Interface (BCI) to train youngsters with Autism Spectrum Disorder (ASD) to follow social cues.

Download

The dataset is publicly available for download here (Amaral et al., 2017; Amaral et al., 2018).

Context

This dataset represents the complete EEG recordings of a feasibility clinical trial (clinical-trial ID: NCT02445625 — clinicaltrials.gov) that tested a P300-based Brain Computer Interface to train youngsters with Autism Spectrum Disorder (ASD) to follow social cues (Amaral et. al, 2017; Amaral et al., 2018).

A total of 15 ASD individuals underwent 7 sessions of P300-based BCI joint-attention training, for a total of 105 sessions. The dataset was used for the 2019 IFMBE Scientific Challenge organized during MEDICON 2019 where, in two phases, teams from all over the world tried to achieve the best possible object-detection accuracy based on the P300 signals.

Content

The dataset folder structure is organized by subjects, with a folder for each subject named SBJXX, with XX varying from 01 to 15. Within each subject folder there is a set of folders containing the data from each session, named SYY, with YY varying from 01 to 07. Each session folder contains a separate folder for the training and testing data, named Train and Test, respectively. The structure and the contents of train and test folders of each session are described in

Train Folder:

• trainEvents.txt – One label per line (from 1 to 8), corresponding to the order of the flashed objects.
• trainData.mat – Data from the calibration phase, structured as [channels x epoch x event], epoch the data samples from -200 ms to 1200 ms relative to the event stimulus.
• trainTargets.txt – 1 or 0 per line, indicating if the flashed object was the target or not, respectively.
• trainLabels.txt – Label of the target object per line (from 1 to 8), one for each block.

Test Folder:

• testData.mat – Data from the online phase, in the same structure as the train data.
• testEvents.txt – One label per line (from 1 to 8), corresponding to the order of the flashed objects.
• runsperblock.txt – File containing only one number, corresponding to the number of runs per block used in the online phase (from 3 to 10).

The number of epochs corresponds to # events per run * # runs per block * # blocks. For the training data, it represents 8 events per run * 10 runs per block * 20 blocks = 1600 epochs. As for the test data, since the number of runs varies between sessions, the number of epochs varies in consequence, in a total of 8 events per run * K runs per block * 50 blocks = 400 * K epochs. The channels’ order in the data matrices is C3, Cz, C4, CPz, P3, Pz, P4, POz. The first sample of each epoch corresponds to the time -200 ms relative to the stimulus onset and the last sample to corresponds to the time 1000 ms, with a sampling rate of 250 Hz.

References

Dataset paper:

Simões M, Borra D, Santamaría-Vázquez E, GBT-UPM, Bittencourt-Villalpando M, Krzemiñski D, Miladinovic´ A, NeuralEngineeringGroup, Schmid T, Zhao H, Amaral C, Direito B, Henriques J, Carvalho P and Castelo-Branco M (2020) BCIAUT-P300: A Multi-Session and Multi-Subject Benchmark Dataset on Autism for P300-Based Brain-Computer-Interfaces. Front. Neurosci. 14:568104. DOI: 10.3389/fnins.2020.568104

Clinical trial paper:

Amaral, C., Mouga, S., Simões, M., Pereira, H. C., Bernardino, I., Quental, H., … Castelo-Branco, M. (2018). A Feasibility Clinical Trial to Improve Social Attention in Autistic Spectrum Disorder (ASD) Using a Brain Computer Interface. Frontiers in Neuroscience, 12(July), 1–13. DOI: 10.3389/fnins.2018.00477

Clinical trial technical paper:

Amaral, C., Simões, M. A., Mouga, S., Andrade, J., & Castelo-Branco, M. (2017). A novel Brain Computer Interface for classification of social joint attention in autism and comparison of 3 experimental setups: A feasibility study. Journal of Neuroscience Methods, 290, 105–115. DOI: 10.1016/j.jneumeth.2017.07.029