Performance evaluation of a motor-imagery-based EEG-Brain computer interface using a combined cue with heterogeneous training data in BCI-Naive subjects

Abstract

BackgroundThe subjects in EEG-Brain computer interface (BCI) system experience difficulties when attempting to obtain the consistent performance of the actual movement by motor imagery alone. It is necessary to find the optimal conditions and stimuli combinations that affect the performance factors of the EEG-BCI system to guarantee equipment safety and trust through the performance evaluation of using motor imagery characteristics that can be utilized in the EEG-BCI testing environment.MethodsThe experiment was carried out with 10 experienced subjects and 32 naive subjects on an EEG-BCI system. There were 3 experiments: The experienced homogeneous experiment, the naive homogeneous experiment and the naive heterogeneous experiment. Each experiment was compared in terms of the six audio-visual cue combinations and consisted of 50 trials. The EEG data was classified using the least square linear classifier in case of the naive subjects through the common spatial pattern filter. The accuracy was calculated using the training and test data set. The p-value of the accuracy was obtained through the statistical significance test.ResultsIn the case in which a naive subject was trained by a heterogeneous combined cue and tested by a visual cue, the result was not only the highest accuracy (p < 0.05) but also stable performance in all experiments.ConclusionsWe propose the use of this measuring methodology of a heterogeneous combined cue for training data and a visual cue for test data by the typical EEG-BCI algorithm on the EEG-BCI system to achieve effectiveness in terms of consistence, stability, cost, time, and resources management without the need for a trial and error process.

Event-related potentials measurement of perception to 3D motion in depth

“…Then, electroencephalography (EEG), recording and measuring electrical signals in line with the neural activities, has been extensively adopted by researchers as a non-invasive approach to exploring human brain's activities [15]. And the Event-related potentials (ERPs), extracted from EEG signals through segmenting, overlapping and averaging, is the measure towards electrical activities highly relating to a specific event [16][17][18][19][20][21][22][23][24].…”

Quantum neural network based surface EMG signal filtering for control of robotic hand

“…The order in which the cues (trials) were presented to the user was generated randomly. In addition, the time between any two trials was also randomized in a range of 0.5 to 2.5 s. Randomness of presenting time of the heterogeneous cue raises the concentration power of the subjects; this is thought to be the main cause of the consistency in performance in similar investigations [22]. In the evaluation phase, feedback to the subjects was given in the form of a number representing the requested cue/class and a number representing the actual finger(s) detection (class) by the classifier.…”

Evaluation of Various Strategies to Improve the Training of a Brain Computer Interface System

“…• Galan et al [10] controlled a virtual wheelchair using a BCI in which the patient has to follow a specific route in a graphic environment • More recent advances include the robotic prosthesis with seven degrees of freedom reported by Collinger et al [11], by the use of a Brain-Machine Interface (BMI), which is similar to a BCI but having the signals from the brain received directly by a machine instead of the computer Another kind of works have been develop to improve the response of the BCI-based system those reported by Choi et al [12] who combined some audio-visual cues so as to improve the training of a electroencephalogram (EEG)-BCI based system; and the one presented by Jin et al [13] where some strategies were applied in order to increase the accuracy and diminish the calibration time for potential (ERP)-based BCIs.…”

“…ther kind of works have been developed in order based systems, as reported by Choi et al [12] who combined some the training of an BCI based system; and the presented by Jin et al [13] After following the instructions given by the to train the system with is the detection suite that [14], we tried to didn't respond most of the times. results, we implemented some the correspondence between the user's will and the actions executed by the mechatronic device using the Cognitive Suite of the SDK.…”