Independent Vector Analysis for Feature Extraction in Motor Imagery Classification

• Published in: Sensors (Basel, Switzerland) Vol. 24; no. 16; p. 5428
• Main Authors: Moraes, Caroline Pires Alavez, dos Santos, Lucas Heck, Fantinato, Denis Gustavo, Neves, Aline, Adali, Tülay
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 22.08.2024; MDPI
• Subjects: Accuracy; Algorithms; autoregressive model; Brain-Computer Interfaces; brain–computer interface; Classification; Datasets; Dependence; electroencephalogram; Electroencephalography; Electroencephalography - methods; Entropy; Humans; Imagination - physiology; independent vector analysis; Methods; motor imagery; Movement - physiology; Signal Processing, Computer-Assisted; Support Vector Machine; Wavelet transforms; brain–computer interface; motor imagery; independent vector analysis; electroencephalogram; autoregressive model
• DOI: 10.3390/s24165428
• ISSN: 1424-8220

Independent vector analysis (IVA) can be viewed as an extension of independent component analysis (ICA) to multiple datasets. It exploits the statistical dependency between different datasets through mutual information. In the context of motor imagery classification based on electroencephalogram (EEG) signals for the brain–computer interface (BCI), several methods have been proposed to extract features efficiently, mainly based on common spatial patterns, filter banks, and deep learning. However, most methods use only one dataset at a time, which may not be sufficient for dealing with a multi-source retrieving problem in certain scenarios. From this perspective, this paper proposes an original approach for feature extraction through multiple datasets based on IVA to improve the classification of EEG-based motor imagery movements. The IVA components were used as features to classify imagined movements using consolidated classifiers (support vector machines and K-nearest neighbors) and deep classifiers (EEGNet and EEGInception). The results show an interesting performance concerning the clustering of MI-based BCI patients, and the proposed method reached an average accuracy of 86.7%.