Comparison of machine learning methods for two class motor imagery tasks using EEG in brain-computer interface

EEG signals are obtained from an EEG device after recording the user's brain signals. EEG signals can be generated by the user after performing motor movements or imagery tasks. Motor Imagery (MI) is the task of imagining motor movements that resemble the original motor movements. Brain Computer Interface (BCI) bridges interactions between users and applications in performing tasks. Brain Computer Interface (BCI) Competition IV 2a was used in this study. A fully automated correction method of EOG artifacts in EEG recordings was applied in order to remove artifacts and Common Spatial Pattern (CSP) to get features that can distinguish motor imagery tasks. In this study, a comparative studies between two deep learning methods was explored, namely Deep Belief Network (DBN) and Long Short Term Memory (LSTM). Usability of both deep learning methods was evaluated using the BCI Competition IV-2a dataset. The experimental results of these two deep learning methods show average accuracy of 50.35% for DBN and 49.65% for LSTM.

Download Full-text

Review of solutions for the application of example of machine learning methods for Motor Imagery in correlation with Brain-Computer Interfaces

PRZEGLĄD ELEKTROTECHNICZNY ◽

10.15199/48.2021.11.20 ◽

2021 ◽

Vol 1 (11) ◽

pp. 113-118

Author(s):

Szczepan PASZKIEL

Keyword(s):

Machine Learning ◽

Motor Imagery ◽

Brain Computer Interfaces ◽

Learning Methods ◽

Machine Learning Methods ◽

Computer Interfaces

Download Full-text

Comparison of Machine Learning Approaches for Motor Imagery Based Optical Brain Computer Interface

Advances in Neuroergonomics and Cognitive Engineering - Advances in Intelligent Systems and Computing ◽

10.1007/978-3-319-94866-9_12 ◽

2018 ◽

pp. 124-134 ◽

Cited By ~ 1

Author(s):

Lei Wang ◽

Adrian Curtin ◽

Hasan Ayaz

Keyword(s):

Machine Learning ◽

Motor Imagery ◽

Brain Computer Interface ◽

Computer Interface ◽

Learning Approaches

Download Full-text

Motor imagery classification in Brain computer interface (BCI) based on EEG signal by using machine learning technique

Bulletin of Electrical Engineering and Informatics ◽

10.11591/eei.v8i1.1402 ◽

2019 ◽

Vol 8 (1) ◽

pp. 269-275 ◽

Cited By ~ 2

Author(s):

N. E. Md Isa ◽

A. Amir ◽

M. Z. Ilyas ◽

M. S. Razalli

Keyword(s):

Machine Learning ◽

Feature Extraction ◽

Logistic Regression ◽

Motor Imagery ◽

Naive Bayes ◽

Brain Computer Interface ◽

Naïve Bayes ◽

Computer Interface ◽

Machine Learning Technique ◽

Learning Technique

This paper focuses on classification of motor imagery in Brain Computer Interface (BCI) by using classifiers from machine learning technique. The BCI system consists of two main steps which are feature extraction and classification. The Fast Fourier Transform (FFT) features is extracted from the electroencephalography (EEG) signals to transform the signals into frequency domain. Due to the high dimensionality of data resulting from the feature extraction stage, the Linear Discriminant Analysis (LDA) is used to minimize the number of dimension by finding the feature subspace that optimizes class separability. Five classifiers: Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Naïve Bayes, Decision Tree and Logistic Regression are used in the study. The performance was tested by using Dataset 1 from BCI Competition IV which consists of imaginary hand and foot movement EEG data. As a result, SVM, Logistic Regression and Naïve Bayes classifier achieved the highest accuracy with 89.09% in AUC measurement.

Download Full-text

The Ensemble Machine Learning-Based Classification of Motor Imagery Tasks in Brain-Computer Interface

Journal of Healthcare Engineering ◽

10.1155/2021/1970769 ◽

2021 ◽

Vol 2021 ◽

pp. 1-12

Author(s):

Abdulhamit Subasi ◽

Saeed Mian Qaisar

Keyword(s):

Machine Learning ◽

Motor Imagery ◽

Brain Computer Interface ◽

Wavelet Packet ◽

Computer Interface ◽

Learning Approach ◽

Wavelet Packet Decomposition ◽

Statistical Features ◽

Ensemble Machine Learning

The Brain-Computer Interface (BCI) permits persons with impairments to interact with the real world without using the neuromuscular pathways. BCIs are based on artificial intelligence piloted systems. They collect brain activity patterns linked to the mental process and transform them into commands for actuators. The potential application of BCI systems is in the rehabilitation centres. In this context, a novel method is devised for automated identification of the Motor Imagery (MI) tasks. The contribution is an effective hybridization of the Multiscale Principal Component Analysis (MSPCA), Wavelet Packet Decomposition (WPD), statistical features extraction from subbands, and ensemble learning-based classifiers for categorization of the MI tasks. The intended electroencephalogram (EEG) signals are segmented and denoised. The denoising is achieved with a Daubechies algorithm-based wavelet transform (WT) incorporated in the MSPCA. The WT with the 5th level of decomposition is used. Onward, the Wavelet Packet Decomposition (WPD), with the 4th level of decomposition, is used for subbands formation. The statistical features are selected from each subband, namely, mean absolute value, average power, standard deviation, skewness, and kurtosis. Also, ratios of absolute mean values of adjacent subbands are computed and concatenated with other extracted features. Finally, the ensemble machine learning approach is used for the classification of MI tasks. The usefulness is evaluated by using the BCI competition III, MI dataset IVa. Results revealed that the suggested ensemble learning approach yields the highest classification accuracies of 98.69% and 94.83%, respectively, for the cases of subject-dependent and subject-independent problems.

Download Full-text

Machine Learning Using Brain Computer Interface System

Volume 5: Biomedical and Biotechnology ◽

10.1115/imece2020-23394 ◽

2020 ◽

Author(s):

Kevin Matsuno ◽

Vidya K. Nandikolla

Keyword(s):

Machine Learning ◽

Motor Imagery ◽

Brain Computer Interface ◽

Hand Movement ◽

Electrical Potential ◽

Machine Learning Algorithms ◽

Computer Interface ◽

Brain Waves ◽

Right Hand ◽

Robotic Application

Abstract With commercially available hardware and supporting software, different electrical potential brain waves are measured via a headset with a collection of electrodes. Out of the different types of brain signals, the proposed brain-computer interface (BCI) controller utilizes non-task related signals, i.e. squeezing left/right hand or tapping left/right foot, due to their responsive behavior and general signal feature similarity among patients. In addition, motor imagery related signals, such as imagining left/right foot or hand movement are also examined. The main goal of the paper is to demonstrate the performance of machine learning algorithms based on classification accuracy. The performances are evaluated on BCI dataset of three male subjects to extract the most significant features. Each subject undergoes a 30-minute session composed of four experiments: two non-task related signals and two motor imagery signals. Each experiment records fifteen trials of two classes (i.e. left/right hand movement). The raw data is then pre-processed using a MatLab plugin, EEGLAB, where standard processes of cleaning and epoching the signals is performed. The paper discusses machine learning for robotic application and the common flaws when validating machine learning methods in the context of BCI to provide a brief overview on biologically (using brain waves) controlled devices.

Download Full-text