This paper proposes the design and the validation through in-vivo measurements, of an innovative machine learning (ML) approach for a synchronous Brain Computer Interface (BCI). The here-proposed system analyzes EEG signals from 8 wireless smart electrodes, placed in motor, and sensory-motor cortex area. For its functioning, the BCI exploits a specific brain activity patterns (BAP) elicited during the measurements by using clinical-inspired stimulation protocol that is suitable for the evocation of the Movement-Related Cortical Potentials (MRCPs). The proposed BCI analyzes the EEGs through symbolization-based algorithm: the Local Binary Patterning, which – due to its end-to-end binary nature - strongly reduces the computational complexity of the features extraction (FE) and real-time classification stages. As last step, the user intentions discrimination is entrusted to a weighted Support Vector Machine (wSVM) with linear kernel. The data have been collected from 3 subjects (aged 26 ± 1), creating an overall dataset that consists of 391 ± 106 observations per participant. The in-vivo real-time validation showed an intention recognition accuracy of 85.61 ± 1.19%. The overall computing chain requests, on average, just 3 ms beyond the storage time.

Novel Synchronous Brain Computer Interface Based on 2-D EEG Local Binary Patterning

De Venuto, Daniela;Mezzina, Giovanni
2020

Abstract

This paper proposes the design and the validation through in-vivo measurements, of an innovative machine learning (ML) approach for a synchronous Brain Computer Interface (BCI). The here-proposed system analyzes EEG signals from 8 wireless smart electrodes, placed in motor, and sensory-motor cortex area. For its functioning, the BCI exploits a specific brain activity patterns (BAP) elicited during the measurements by using clinical-inspired stimulation protocol that is suitable for the evocation of the Movement-Related Cortical Potentials (MRCPs). The proposed BCI analyzes the EEGs through symbolization-based algorithm: the Local Binary Patterning, which – due to its end-to-end binary nature - strongly reduces the computational complexity of the features extraction (FE) and real-time classification stages. As last step, the user intentions discrimination is entrusted to a weighted Support Vector Machine (wSVM) with linear kernel. The data have been collected from 3 subjects (aged 26 ± 1), creating an overall dataset that consists of 391 ± 106 observations per participant. The in-vivo real-time validation showed an intention recognition accuracy of 85.61 ± 1.19%. The overall computing chain requests, on average, just 3 ms beyond the storage time.
Intelligent Systems and Applications : Proceedings of the 2019 Intelligent Systems Conference (IntelliSys). Volume 2
978-3-030-29512-7
Springer
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11589/181124
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • Scopus 1
  • ???jsp.display-item.citation.isi??? 0
social impact