DICIS EEGFMMI A dataset

Published: 11 March 2021| Version 1 | DOI: 10.17632/2sdzh8dgvt.1
Contributors:
Tat´y Mwata,
,

Description

The dataset was created for the Motor Imagery classification by using an EMOTIV EPOC+ headset, to control a hexapod robot. Four test subjects between 23 and 36 years old, were trained and supervised to collect signals during several experimental tasks lasting three seconds each. According to the given task, subjects were instructed to stay still during the capture and invited to imagine closing and opening the right or the left fist focused on a stimulus video. The stimulus video of the fist closing-opening movements was played on the screen according to the temporal task sequence. The total trial duration was established at 19 s, where 6 s were used for the related MI tasks: the left fist as Task 1 (3 s), and the right fist as Task 2 (3 s). The neutral or reference action is taken as the final pause (3 s) to have an equal number of samples per class. Therefore, the developed dataset consists of 2,400 trials performed by four subjects (600 trials from each subject), representing 2,400×19 s (12.67 hours) of data capture. For each session duration, only signals of 3 s corresponding to Task1 (left fist MI), 3 s toTask2 (right fist MI), and 3 s to neutral action have been gathered to build the dataset. The Data folder contains EEG data corresponding to 4 subjects. Each subject folder contains 3 CSV files (one for each task). Each CSV file is organized as follows: the header contains the sensor names and the following lines contain the signals corresponding to 200 trials (each trial contains 384 samples); therefore, the total number of samples is 76,800 (384 x 200 = 76,800 samples).

Files

Steps to reproduce

1. A subject comfortably seated and wearing an EMOTIV EPOC+ headset is instructed to stay still during the capture and invited to imagine closing and opening the right or the left fist focused on a stimulus video. 2. In the capture sequence, the first five-seconds serve to prepare the test subject ending this phase with an audible Beep1. 3. Task 1, related to the left fist MI task, is executed ending with a Beep2 tone. 4. A pause of 3 seconds is executed. Beep3 triggers the end of this static period and starting a second preparation phase of 2s. 5. Beep4 starts Task 2, related to the right fist MI task, ending with Beep5. For each session duration, only signals of 3 s correspondingtoTask1(left fist MI), 3 s toTask2(right fist MI), and 3 s to neutral action have been gathered to build the dataset (see CaptureSequence.pdf).

Institutions

Universidad de Guanajuato - Campus Irapuato-Salamanca

Categories

Signal Processing, Motor Cortex, Electroencephalography, Brain-Computer Interface, Recurrent Neural Network, Deep Learning

License