Dendrite morphological neural networks for motor task recognition from electroencephalographic signals

• Published in: Biomedical signal processing and control Vol. 44; pp. 12 - 24
• Main Authors: Antelis, Javier M., Gudiño-Mendoza, Berenice, Falcón, Luis Eduardo, Sanchez-Ante, Gildardo, Sossa, Humberto
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.07.2018
• Subjects: Brain–computer interface; Classification accuracy; Electroencephalogram; Morphological; Motor task; Neural network; Motor task; Neural network; Electroencephalogram; Classification accuracy; Morphological; Brain–computer interface
• ISSN: 1746-8094; 1746-8108
• DOI: 10.1016/j.bspc.2018.03.010

[Display omitted] •DMNN are proposed to recognize motor tasks from EEG signals.•A systematic comparison was performed with classical classifications algorithms.•DMNN is a promising technique for the recognition of motor tasks from EEG signals.•Recognition of the intention to move is important to reduce response time in BCIs. Brain–computer interfaces (BCI) rely on classification algorithms to detect the patterns of the brain signals that encode the mental task performed by the user. Therefore, robust and reliable classification techniques should be developed and evaluated to recognize the user's mental task with high accuracy. This paper proposes the use of the novel dendrite morphological neural networks (DMNN) for the recognition of voluntary movements from electroencephalographic (EEG) signals. This technique was evaluated with two studies. The first aimed to evaluate the performance of DMNN in the recognition of motor execution and motor imagery tasks and to carry out a systematic comparison with support vector machine (SVM) and linear discriminant analysis (LDA) which are the two classifiers mostly used in BCI systems. EEG signals from twelve healthy students were recorded during a cue-based hand motor execution and imagery experiment. The results showed that DMNN provided decoding accuracies of 80% for motor execution and 77% for motor imagery, which were significantly different than the chance level (p < 0.05, Wilcoxon signed-rank test) and higher when compared with classifiers commonly used in BCI. The second study aimed to employ the DMNN to recognize the intention of movement. To this end, EEG signals were recorded from eighteen healthy subjects performing self-paced reaching movements and several classification scenarios were evaluated. The results showed that DMNN provided decoding accuracies above chance level, whereby, it is able to detect a movement prior its execution. On the basis of these results, DMNN is a powerful promising classification technique that can be used to enhance performance in the recognition of motor tasks for BCI systems based on electroencephalographic signals.