Embedded Brain Machine Interface based on motor imagery paradigm to control prosthetic hand

Kevin Acuna; Erick Carranza Urquizo; David Achanccaray

Embedded Brain Machine Interface based on motor imagery paradigm to control prosthetic hand

Kevin Acuna
, Erick Carranza Urquizo
, David Achanccaray

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

13 Scopus citations

Abstract

Brain Machine Interfaces (BMI) have been developed as an alternative way to decode brain signals into control commands and communication devices. A typical BMI uses a computer to process EEG signals; however, current embedded PCs have enough computational resources for fully embedded BMI systems. In this work, the performance of the Odroid-xu4 embedded PC is evaluated as a processing and control device for BMI based on a 2-class motor imagery paradigm. Results show the best accuracy (82.1%) using SVM classifier and minimal processing times (0.11s) on the embedded device, which allows the development of a portable, low cost and trustworthy system.

Original language	Spanish
Title of host publication	Proceedings of the 2016 IEEE ANDESCON, ANDESCON 2016
State	Published - 27 Jan 2017

Cite this

@inproceedings{8d285ed228a14b52bc5f39d5d85e3cd2,

title = "Embedded Brain Machine Interface based on motor imagery paradigm to control prosthetic hand",

abstract = "Brain Machine Interfaces (BMI) have been developed as an alternative way to decode brain signals into control commands and communication devices. A typical BMI uses a computer to process EEG signals; however, current embedded PCs have enough computational resources for fully embedded BMI systems. In this work, the performance of the Odroid-xu4 embedded PC is evaluated as a processing and control device for BMI based on a 2-class motor imagery paradigm. Results show the best accuracy (82.1\%) using SVM classifier and minimal processing times (0.11s) on the embedded device, which allows the development of a portable, low cost and trustworthy system.",

author = "Kevin Acuna and Urquizo, \{Erick Carranza\} and David Achanccaray",

year = "2017",

month = jan,

day = "27",

language = "Espa{\~n}ol",

booktitle = "Proceedings of the 2016 IEEE ANDESCON, ANDESCON 2016",

}

TY - GEN

T1 - Embedded Brain Machine Interface based on motor imagery paradigm to control prosthetic hand

AU - Acuna, Kevin

AU - Urquizo, Erick Carranza

AU - Achanccaray, David

PY - 2017/1/27

Y1 - 2017/1/27

N2 - Brain Machine Interfaces (BMI) have been developed as an alternative way to decode brain signals into control commands and communication devices. A typical BMI uses a computer to process EEG signals; however, current embedded PCs have enough computational resources for fully embedded BMI systems. In this work, the performance of the Odroid-xu4 embedded PC is evaluated as a processing and control device for BMI based on a 2-class motor imagery paradigm. Results show the best accuracy (82.1%) using SVM classifier and minimal processing times (0.11s) on the embedded device, which allows the development of a portable, low cost and trustworthy system.

AB - Brain Machine Interfaces (BMI) have been developed as an alternative way to decode brain signals into control commands and communication devices. A typical BMI uses a computer to process EEG signals; however, current embedded PCs have enough computational resources for fully embedded BMI systems. In this work, the performance of the Odroid-xu4 embedded PC is evaluated as a processing and control device for BMI based on a 2-class motor imagery paradigm. Results show the best accuracy (82.1%) using SVM classifier and minimal processing times (0.11s) on the embedded device, which allows the development of a portable, low cost and trustworthy system.

M3 - Contribución a la conferencia

BT - Proceedings of the 2016 IEEE ANDESCON, ANDESCON 2016

ER -