Volume 7, Issue 1 (Winter 2016 -- 2016)                   BCN 2016, 7(1): 13-20 | Back to browse issues page

XML Print

Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

Resalat S N, Saba V. A Study of Various Feature Extraction Methods on a Motor Imagery Based Brain Computer Interface System. BCN. 2016; 7 (1) :13-20
URL: http://bcn.iums.ac.ir/article-1-384-en.html
1- Organization
2- Radiation Research Center, Faculty of Paramedicine, AJA University of Medical Sciences, Tehran, Iran.

Introduction: Brain Computer Interface (BCI) systems based on Movement Imagination (MI) are widely used in recent decades. Separate feature extraction methods are employed in the MI data sets and classified in Virtual Reality (VR) environments for real-time applications.
Methods: This study applied wide variety of features on the recorded data using Linear Discriminant Analysis (LDA) classifier to select the best feature sets in the offline mode. The data set was recorded in 3-class tasks of the left hand, the right hand, and the foot motor imagery.
Results: The experimental results showed that Auto-Regressive (AR), Mean Absolute Value (MAV), and Band Power (BP) features have higher accuracy values,75% more than those for the other features.
Discussion: These features were selected for the designed real-time navigation. The corresponding results revealed the subject-specific nature of the MI-based BCI system however, the Power Spectral Density (PSD) based &alpha-BP feature had the highest averaged accuracy.

Type of Study: Original | Subject: Cognitive Neuroscience
Received: 2015/02/28 | Accepted: 2015/06/21 | Published: 2016/01/1

Add your comments about this article : Your username or Email:

Send email to the article author

© 2019 All Rights Reserved | Basic and Clinical Neuroscience

Designed & Developed by : Yektaweb